diff --git a/contrib/templates/llmops-basic/README.md b/contrib/templates/llmops-basic/README.md
new file mode 100644
index 0000000..eb43a3b
--- /dev/null
+++ b/contrib/templates/llmops-basic/README.md
@@ -0,0 +1,10 @@
+# LLMOps basic template
+
+This template introduces a base structure for organizing LLMOps projects
+using DABs, Unity Catalog and MLflow.
+
+Install it using
+
+```
+databricks bundle init https://github.com/databricks/bundle-examples --template-dir contrib/templates/llmops-basic
+```
\ No newline at end of file
diff --git a/contrib/templates/llmops-basic/databricks_template_schema.json b/contrib/templates/llmops-basic/databricks_template_schema.json
new file mode 100644
index 0000000..7162166
--- /dev/null
+++ b/contrib/templates/llmops-basic/databricks_template_schema.json
@@ -0,0 +1,35 @@
+{
+  "properties": {
+    "project_name": {
+      "type": "string",
+      "default": "llmops_basic",
+      "description": "Name of the LLMOps project (use underscores, no hyphens)",
+      "order": 1
+    },
+    "catalog_name_dev": {
+      "type": "string",
+      "default": "dev_catalog",
+      "description": "Name of the Unity Catalog for development environment",
+      "order": 2
+    },
+    "catalog_name_prod": {
+      "type": "string",
+      "default": "prod_catalog",
+      "description": "Name of the Unity Catalog for production environment",
+      "order": 3
+    },
+    "workspace_host_dev": {
+      "type": "string",
+      "default": "https://your-workspace.azuredatabricks.net/",
+      "description": "Databricks workspace URL for development",
+      "order": 4
+    },
+    "workspace_host_prod": {
+      "type": "string",
+      "default": "https://your-workspace.azuredatabricks.net/",
+      "description": "Databricks workspace URL for production",
+      "order": 5
+    }
+  }
+}
+
diff --git a/contrib/templates/llmops-basic/template/{{.project_name}}/.gitignore b/contrib/templates/llmops-basic/template/{{.project_name}}/.gitignore
new file mode 100644
index 0000000..f3b15af
--- /dev/null
+++ b/contrib/templates/llmops-basic/template/{{.project_name}}/.gitignore
@@ -0,0 +1,22 @@
+.databricks/
+.bundle/
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+.Python
+venv/
+env/
+ENV/
+.ipynb_checkpoints/
+*.ipynb_checkpoints
+.vscode/
+.idea/
+*.swp
+*.swo
+*~
+.DS_Store
+Thumbs.db
+mlruns/
+*.log
+
diff --git a/contrib/templates/llmops-basic/template/{{.project_name}}/README.md.tmpl b/contrib/templates/llmops-basic/template/{{.project_name}}/README.md.tmpl
new file mode 100644
index 0000000..cac94b8
--- /dev/null
+++ b/contrib/templates/llmops-basic/template/{{.project_name}}/README.md.tmpl
@@ -0,0 +1,44 @@
+# {{.project_name}}
+
+End-to-end LLMOps project for sentiment analysis of reviews using Databricks.
+
+## Overview
+
+This example demonstrates a complete LLMOps pipeline for building, evaluating, and deploying a sentiment analysis model.
+
+**Pipeline stages:**
+- Data preparation
+- Model build and evaluation with MLflow
+- Model deployment to serving endpoint
+- Batch inference
+
+## Requirements
+
+- Databricks CLI (v0.218.0+)
+- Unity Catalog enabled
+- Required permissions on the catalog:
+  - `USE CATALOG` - to access the catalog
+  - `CREATE SCHEMA` - to create schemas
+
+## Quick Start
+
+**Deploy:**
+```bash
+databricks bundle deploy -t dev
+```
+
+**Run pipeline:**
+```bash
+databricks bundle run model_preprocessing -t dev
+databricks bundle run model_build_evaluation -t dev
+databricks bundle run model_endpoint_deploy -t dev
+databricks bundle run model_inference -t dev
+```
+
+## Configuration
+
+- **Dev Catalog**: `{{.catalog_name_dev}}`
+- **Prod Catalog**: `{{.catalog_name_prod}}`
+
+Edit `databricks.yml` to customize catalog names, schema name, model name, and experiment settings.
+
diff --git a/contrib/templates/llmops-basic/template/{{.project_name}}/databricks.yml.tmpl b/contrib/templates/llmops-basic/template/{{.project_name}}/databricks.yml.tmpl
new file mode 100644
index 0000000..c521ed9
--- /dev/null
+++ b/contrib/templates/llmops-basic/template/{{.project_name}}/databricks.yml.tmpl
@@ -0,0 +1,44 @@
+# This is a Databricks Asset Bundle for {{.project_name}}.
+bundle:
+  name: "{{.project_name}}"
+
+variables:
+  catalog_name:
+    description: "Name of the UC catalog to use"
+    default: "default_catalog"
+  schema_name:
+    description: "Name of the UC schema to use"
+    default: "sentiment_agent_project"
+  model_name:
+    description: "Name of the UC model to use"
+    default: "sentiment_agent"
+  experiment_name:
+    description: "Name of experiment"
+    default: "/Users/${workspace.current_user.userName}/${bundle.target}_sentiment_agent"
+
+include:
+  - resources/*.yml
+
+targets:
+  dev:
+    mode: development
+    default: true
+    workspace:
+      host: {{.workspace_host_dev}}
+    variables:
+      catalog_name: {{.catalog_name_dev}}
+
+  prod:
+    mode: production
+    workspace:
+      host: {{.workspace_host_prod}}
+      root_path: /Shared/.bundle/prod/${bundle.name}
+    variables:
+      catalog_name: {{.catalog_name_prod}}
+    {{- if not is_service_principal}}
+    run_as:
+      # This runs as {{user_name}} in production. Alternatively,
+      # a service principal could be used here using service_principal_name
+      # (see Databricks documentation).
+      user_name: {{user_name}}
+    {{end -}}
diff --git a/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/model_build.ipynb b/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/model_build.ipynb
new file mode 100644
index 0000000..8ea1c35
--- /dev/null
+++ b/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/model_build.ipynb
@@ -0,0 +1,276 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "64ed089a-931a-4123-afdc-a2c38390774c",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "%pip install -U -qqqq mlflow databricks-openai databricks-agents\n",
+    "dbutils.library.restartPython()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "7f453049-63d9-46b2-8421-92f8c6a5442a",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "dbutils.widgets.text(\"catalog_name\", \"dev_catalog\")\n",
+    "dbutils.widgets.text(\"schema_name\", \"default\")\n",
+    "dbutils.widgets.text(\"model_name\", \"quickstart_agent\")\n",
+    "\n",
+    "\n",
+    "catalog_name = dbutils.widgets.get(\"catalog_name\")\n",
+    "schema_name = dbutils.widgets.get(\"schema_name\")\n",
+    "model_name = dbutils.widgets.get(\"model_name\")\n",
+    "dbutils.widgets.text(\"experiment_name\", f\"/Users/{dbutils.notebook.entry_point.getDbutils().notebook().getContext().userName().get()}/{model_name}_{catalog_name}\")\n",
+    "experiment_name = dbutils.widgets.get(\"experiment_name\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "114153c0-b167-451b-a2ac-6d4f48faeaec",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "import mlflow\n",
+    "\n",
+    "registered_model_name = f\"{catalog_name}.{schema_name}.{model_name}\"\n",
+    "\n",
+    "mlflow.set_experiment(experiment_name)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "154fbd28-5ad2-4a51-8ccf-fa850dffc5bd",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "from mlflow.models.resources import DatabricksFunction, DatabricksServingEndpoint\n",
+    "from pkg_resources import get_distribution\n",
+    "from quickstart_agent import LLM_ENDPOINT_NAME\n",
+    "import datetime\n",
+    "\n",
+    "mlflow.set_registry_uri(\"databricks-uc\")\n",
+    "resources = [\n",
+    "    DatabricksServingEndpoint(endpoint_name=LLM_ENDPOINT_NAME),\n",
+    "]\n",
+    "\n",
+    "timestamp = datetime.datetime.now().strftime(\"%Y%m%d_%H%M%S\")\n",
+    "\n",
+    "with mlflow.start_run(run_name=f\"sentiment_agent_{timestamp}\"):\n",
+    "    logged_model_info = mlflow.pyfunc.log_model(\n",
+    "        artifact_path=\"agent\",\n",
+    "        python_model=\"quickstart_agent.py\",\n",
+    "        extra_pip_requirements=[\n",
+    "            f\"databricks-connect=={get_distribution('databricks-connect').version}\"\n",
+    "        ],\n",
+    "        resources=resources\n",
+    "    )"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "ea348fb4-9a9f-4772-aaf1-65a646ca5739",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "dbutils.jobs.taskValues.set(key = \"logged_run_id\", value = logged_model_info.run_id)"
+   ]
+  }
+ ],
+ "metadata": {
+  "application/vnd.databricks.v1+notebook": {
+   "computePreferences": {
+    "hardware": {
+     "accelerator": null,
+     "gpuPoolId": null,
+     "memory": null
+    }
+   },
+   "dashboards": [],
+   "environmentMetadata": {
+    "base_environment": "",
+    "environment_version": "3"
+   },
+   "inputWidgetPreferences": null,
+   "language": "python",
+   "notebookMetadata": {
+    "pythonIndentUnit": 2
+   },
+   "notebookName": "model_build",
+   "widgets": {
+    "catalog_name": {
+     "currentValue": "dev_catalog",
+     "nuid": "fd954ea2-6905-4526-b686-38b43c5b6d97",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "dev_catalog",
+      "label": null,
+      "name": "catalog_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "dev_catalog",
+      "label": null,
+      "name": "catalog_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "experiment_name": {
+     "currentValue": "",
+     "nuid": "818e024c-c654-4a9c-9f36-7e595b9a9e5f",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "tee",
+      "label": null,
+      "name": "experiment_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "tee",
+      "label": null,
+      "name": "experiment_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "model_name": {
+     "currentValue": "quickstart_agent",
+     "nuid": "93dd77ea-7b35-410c-af71-589d260ae712",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "quickstart_agent",
+      "label": null,
+      "name": "model_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "quickstart_agent",
+      "label": null,
+      "name": "model_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "schema_name": {
+     "currentValue": "default",
+     "nuid": "88bd61ef-aa5c-42a2-ad26-f9079c1afc5f",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "default",
+      "label": null,
+      "name": "schema_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "default",
+      "label": null,
+      "name": "schema_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    }
+   }
+  },
+  "language_info": {
+   "name": "python"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
+}
diff --git a/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/model_endpoint_deploy.ipynb b/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/model_endpoint_deploy.ipynb
new file mode 100644
index 0000000..2f18515
--- /dev/null
+++ b/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/model_endpoint_deploy.ipynb
@@ -0,0 +1,118 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {},
+     "inputWidgets": {},
+     "nuid": "28525745-4fb6-4667-92f1-1e0e3fa68da1",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "%pip install -U -qqqq mlflow databricks-openai databricks-agents\n",
+    "dbutils.library.restartPython()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {},
+     "inputWidgets": {},
+     "nuid": "b2519e1e-ef7e-439c-ab55-5c17e77c8b25",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "dbutils.widgets.text(\"catalog_name\", \"dev_catalog\")\n",
+    "dbutils.widgets.text(\"schema_name\", \"default\")\n",
+    "dbutils.widgets.text(\"model_name\", \"quickstart_agent\")\n",
+    "\n",
+    "catalog_name = dbutils.widgets.get(\"catalog_name\")\n",
+    "schema_name = dbutils.widgets.get(\"schema_name\")\n",
+    "model_name = dbutils.widgets.get(\"model_name\")\n",
+    "dbutils.widgets.text(\"experiment_name\", f\"/Users/{dbutils.notebook.entry_point.getDbutils().notebook().getContext().userName().get()}/{model_name}_{catalog_name}\")\n",
+    "experiment_name = dbutils.widgets.get(\"experiment_name\")\n",
+    "\n",
+    "import mlflow\n",
+    "\n",
+    "mlflow.set_experiment(experiment_name)\n",
+    "\n",
+    "registered_model_name = f\"{catalog_name}.{schema_name}.{model_name}\""
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {},
+     "inputWidgets": {},
+     "nuid": "b2a5f820-740b-4e9d-8496-4b54c3350156",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "#https://docs.databricks.com/aws/en/generative-ai/agent-framework/deploy-agent#deploy-an-agent-using-deploy\n",
+    "\n",
+    "from mlflow import MlflowClient\n",
+    "from databricks.agents import get_deployments, delete_deployment\n",
+    "from databricks import agents\n",
+    "\n",
+    "client = MlflowClient()\n",
+    "\n",
+    "registered_model_version = client.get_model_version_by_alias(registered_model_name, \"Champion\")\n",
+    "\n",
+    "deployments = get_deployments(model_name=registered_model_name)\n",
+    "\n",
+    "latest_version = registered_model_version.version\n",
+    "\n",
+    "for deployment in deployments:\n",
+    "    if deployment.model_name == registered_model_name:\n",
+    "        print(f\"Deleting deployment: model={deployment.model_name}, version={deployment.model_version}\")\n",
+    "        delete_deployment(model_name=deployment.model_name, model_version=deployment.model_version)\n",
+    "\n",
+    "\n",
+    "deployment_info = agents.deploy(\n",
+    "    model_name=registered_model_name,\n",
+    "    model_version=registered_model_version.version,\n",
+    ")"
+   ]
+  }
+ ],
+ "metadata": {
+  "application/vnd.databricks.v1+notebook": {
+   "computePreferences": null,
+   "dashboards": [],
+   "environmentMetadata": {
+    "base_environment": "",
+    "environment_version": "3"
+   },
+   "inputWidgetPreferences": null,
+   "language": "python",
+   "notebookMetadata": {
+    "pythonIndentUnit": 2
+   },
+   "notebookName": "model_endpoint_deploy",
+   "widgets": {}
+  },
+  "language_info": {
+   "name": "python"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
+}
diff --git a/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/model_evaluation.ipynb b/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/model_evaluation.ipynb
new file mode 100644
index 0000000..5d08f41
--- /dev/null
+++ b/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/model_evaluation.ipynb
@@ -0,0 +1,396 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "be63921b-4820-4872-a708-bf5722e0f8e7",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "%pip install -U -qqqq mlflow databricks-openai databricks-agents threadpoolctl==3.1.0\n",
+    "dbutils.library.restartPython()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "8af65798-0acc-4263-bc35-9212653d3987",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "dbutils.widgets.text(\"catalog_name\", \"dev_catalog\")\n",
+    "dbutils.widgets.text(\"schema_name\", \"default\")\n",
+    "dbutils.widgets.text(\"model_name\", \"quickstart_agent\")\n",
+    "dbutils.widgets.text(\"logged_run_id\", \"None\")\n",
+    "\n",
+    "catalog_name = dbutils.widgets.get(\"catalog_name\")\n",
+    "schema_name = dbutils.widgets.get(\"schema_name\")\n",
+    "model_name = dbutils.widgets.get(\"model_name\")\n",
+    "logged_run_id = dbutils.widgets.get(\"logged_run_id\")\n",
+    "\n",
+    "dbutils.widgets.text(\"experiment_name\", f\"/Users/{dbutils.notebook.entry_point.getDbutils().notebook().getContext().userName().get()}/{model_name}_{catalog_name}\")\n",
+    "experiment_name = dbutils.widgets.get(\"experiment_name\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "fec9759b-b50d-4611-be17-d3b51a3041a6",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "import mlflow\n",
+    "import mlflow.genai.datasets\n",
+    "import time\n",
+    "from databricks.connect import DatabricksSession\n",
+    "\n",
+    "registered_model_name = f\"{catalog_name}.{schema_name}.{model_name}\"\n",
+    "model_uri = f\"runs:/{logged_run_id}/agent\"\n",
+    "\n",
+    "mlflow.set_experiment(experiment_name)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "f0a33257-a9da-4f65-8c8a-657612d6d634",
+     "showTitle": false,
+     "tableResultSettingsMap": {
+      "0": {
+       "dataGridStateBlob": "{\"version\":1,\"tableState\":{\"columnPinning\":{\"left\":[\"#row_number#\"],\"right\":[]},\"columnSizing\":{},\"columnVisibility\":{}},\"settings\":{\"columns\":{}},\"syncTimestamp\":1756760635746}",
+       "filterBlob": null,
+       "queryPlanFiltersBlob": null,
+       "tableResultIndex": 0
+      }
+     },
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "query = f\"SELECT * FROM {catalog_name}.{schema_name}.labelled_sentences\"\n",
+    "df = spark.sql(query).toPandas()\n",
+    "display(df)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "0c997521-a041-46ed-8b37-b0b9ddf67df3",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "import mlflow\n",
+    "from mlflow.genai.scorers import Guidelines, Correctness, RelevanceToQuery\n",
+    "import datetime\n",
+    "\n",
+    "eval_data = []\n",
+    "for _, row in df.iterrows():\n",
+    "    eval_item = {\n",
+    "        \"inputs\": {\n",
+    "            \"content\": row[\"review\"]  # This matches the function parameter name\n",
+    "        },\n",
+    "        \"expectations\": {\n",
+    "            \"expected_response\": str(row[\"sentiment\"])  # Adjust column name\n",
+    "        }\n",
+    "    }\n",
+    "    eval_data.append(eval_item)\n",
+    "\n",
+    "mlflow_eval_dataset = None\n",
+    "    \n",
+    "try:\n",
+    "    # Try to get existing dataset\n",
+    "    mlflow_eval_dataset = mlflow.genai.get_dataset(f\"{catalog_name}.{schema_name}.mlflow_eval_dataset\")\n",
+    "    print(f\"✓ Using existing dataset: {catalog_name}.{schema_name}.mlflow_eval_dataset\")\n",
+    "    \n",
+    "except Exception as get_error:\n",
+    "    print(f\"Dataset not found, creating dataset\")\n",
+    "    \n",
+    "    try:\n",
+    "        # Create new dataset\n",
+    "        mlflow_eval_dataset = mlflow.genai.create_dataset(f\"{catalog_name}.{schema_name}.mlflow_eval_dataset\")\n",
+    "        mlflow_eval_dataset.merge_records(eval_data)\n",
+    "        dataset_created = True\n",
+    "        print(f\"✓ Created new dataset: {catalog_name}.{schema_name}.mlflow_eval_dataset\")\n",
+    "        \n",
+    "    except Exception as create_error:\n",
+    "        print(f\"⚠ Could not create dataset {create_error}\")\n",
+    "\n",
+    "if (mlflow_eval_dataset is not None):       \n",
+    "        \n",
+    "    guidelines = {\n",
+    "        \"sentiment_accuracy\": \"Response must correctly identify sentiment\",\n",
+    "        \"clarity\": [\"Response must be clear and concise\"]\n",
+    "    }\n",
+    "\n",
+    "    agent = mlflow.pyfunc.load_model(model_uri)\n",
+    "\n",
+    "    def predict_function(content):\n",
+    "        try:\n",
+    "            messages = [\n",
+    "                {\n",
+    "                    \"role\": \"user\",\n",
+    "                    \"content\": content\n",
+    "                }\n",
+    "            ]\n",
+    "            prediction = agent.predict({\"messages\": messages})\n",
+    "            return prediction\n",
+    "        except Exception as e:\n",
+    "            print(f\"Error in predict_fn: {e}\")\n",
+    "            return {\"response\": \"PREDICTION_ERROR\"}\n",
+    "        \n",
+    "    timestamp = datetime.datetime.now().strftime(\"%Y%m%d_%H%M%S\")\n",
+    "    with mlflow.start_run(run_id=logged_run_id) as run:\n",
+    "        with mlflow.start_run(run_name=f\"sentiment_agent_evaluation_{timestamp}\", nested=True) as eval_run:\n",
+    "            results = mlflow.genai.evaluate(\n",
+    "                data=mlflow_eval_dataset,\n",
+    "                predict_fn=predict_function,\n",
+    "                scorers=[\n",
+    "                    Correctness(),\n",
+    "                    Guidelines(name=\"sentiment_accuracy\", guidelines=guidelines[\"sentiment_accuracy\"]),\n",
+    "                    Guidelines(name=\"clarity\", guidelines=guidelines[\"clarity\"]),\n",
+    "                ],\n",
+    "            )\n",
+    "\n",
+    "    # Print results\n",
+    "    print(\"Evaluation Results:\")\n",
+    "    print(results.metrics)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {},
+     "inputWidgets": {},
+     "nuid": "281e07cf-0163-453a-9256-c508b37cd073",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "from mlflow import MlflowClient\n",
+    "client = MlflowClient()\n",
+    "\n",
+    "registered_model = mlflow.register_model(model_uri, name=registered_model_name)\n",
+    "\n",
+    "if results.metrics['sentiment_accuracy/mean'] > 0.9:\n",
+    "  print(\"Transitioning to champion\")\n",
+    "  client.set_registered_model_alias(registered_model_name, \"Champion\",registered_model.version)"
+   ]
+  }
+ ],
+ "metadata": {
+  "application/vnd.databricks.v1+notebook": {
+   "computePreferences": {
+    "hardware": {
+     "accelerator": null,
+     "gpuPoolId": null,
+     "memory": null
+    }
+   },
+   "dashboards": [],
+   "environmentMetadata": {
+    "base_environment": "",
+    "environment_version": "3"
+   },
+   "inputWidgetPreferences": null,
+   "language": "python",
+   "notebookMetadata": {
+    "pythonIndentUnit": 2
+   },
+   "notebookName": "model_evaluation",
+   "widgets": {
+    "catalog_name": {
+     "currentValue": "dev_catalog",
+     "nuid": "3a964089-b210-4dc6-aa79-c7c287530d31",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "dev_catalog",
+      "label": null,
+      "name": "catalog_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "dev_catalog",
+      "label": null,
+      "name": "catalog_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "experiment_name": {
+     "currentValue": "/Users/user@company.com/quickstart_agent_dev_catalog",
+     "nuid": "d243520b-4f07-4843-b294-0e3b3489d85b",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "/Users/user@company.com/quickstart_agent_dev_catalog",
+      "label": null,
+      "name": "experiment_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "/Users/user@company.com/quickstart_agent_dev_catalog",
+      "label": null,
+      "name": "experiment_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "logged_run_id": {
+     "currentValue": "None",
+     "nuid": "970893b9-068c-4fff-af42-e9b346618f94",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "None",
+      "label": null,
+      "name": "logged_run_id",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "None",
+      "label": null,
+      "name": "logged_run_id",
+      "options": {
+       "autoCreated": false,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "model_name": {
+     "currentValue": "quickstart_agent",
+     "nuid": "06926edb-a96f-4f02-a702-f7c542e5ef3c",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "quickstart_agent",
+      "label": null,
+      "name": "model_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "quickstart_agent",
+      "label": null,
+      "name": "model_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "schema_name": {
+     "currentValue": "default",
+     "nuid": "febc8306-c467-4694-9bfe-6467bb3558a2",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "default",
+      "label": null,
+      "name": "schema_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "default",
+      "label": null,
+      "name": "schema_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    }
+   }
+  },
+  "language_info": {
+   "name": "python"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
+}
diff --git a/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/model_inference.ipynb b/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/model_inference.ipynb
new file mode 100644
index 0000000..e0c4535
--- /dev/null
+++ b/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/model_inference.ipynb
@@ -0,0 +1,243 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "be63921b-4820-4872-a708-bf5722e0f8e7",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "%pip install -U -qqqq mlflow databricks-openai databricks-agents\n",
+    "dbutils.library.restartPython()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "138d2537-42eb-4fae-82b8-e11021a8a85b",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "dbutils.widgets.text(\"catalog_name\", \"dev_catalog\")\n",
+    "dbutils.widgets.text(\"schema_name\", \"default\")\n",
+    "dbutils.widgets.text(\"model_name\", \"quickstart_agent\")\n",
+    "catalog_name = dbutils.widgets.get(\"catalog_name\")\n",
+    "schema_name = dbutils.widgets.get(\"schema_name\")\n",
+    "model_name = dbutils.widgets.get(\"model_name\")\n",
+    "\n",
+    "dbutils.widgets.text(\"experiment_name\", f\"/Users/{dbutils.notebook.entry_point.getDbutils().notebook().getContext().userName().get()}/{model_name}_{catalog_name}\")\n",
+    "experiment_name = dbutils.widgets.get(\"experiment_name\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "fec9759b-b50d-4611-be17-d3b51a3041a6",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "import mlflow\n",
+    "import mlflow.genai.datasets\n",
+    "import time\n",
+    "from databricks.connect import DatabricksSession\n",
+    "\n",
+    "mlflow.set_experiment(experiment_name)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "40bbaf35-a61d-439f-9bf0-b160a96223ab",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "query = f\"SELECT * FROM {catalog_name}.{schema_name}.sentences\"\n",
+    "df = spark.sql(query).toPandas()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "0c997521-a041-46ed-8b37-b0b9ddf67df3",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "\n",
+    "model_uri = f\"models:/{catalog_name}.{schema_name}.{model_name}@champion\"\n",
+    "agent = mlflow.pyfunc.load_model(model_uri)\n",
+    "\n",
+    "def predict_sentiment(review):\n",
+    "    messages = [{\n",
+    "        \"role\": \"user\",\n",
+    "        \"content\": f\"{review}\"\n",
+    "    }]\n",
+    "    return agent.predict({\"messages\": messages})\n",
+    "\n",
+    "df['predictions'] = df['review'].apply(predict_sentiment)\n",
+    "display(df)\n"
+   ]
+  }
+ ],
+ "metadata": {
+  "application/vnd.databricks.v1+notebook": {
+   "computePreferences": {
+    "hardware": {
+     "accelerator": null,
+     "gpuPoolId": null,
+     "memory": null
+    }
+   },
+   "dashboards": [],
+   "environmentMetadata": {
+    "base_environment": "",
+    "environment_version": "3"
+   },
+   "inputWidgetPreferences": null,
+   "language": "python",
+   "notebookMetadata": {
+    "pythonIndentUnit": 2
+   },
+   "notebookName": "model_inference",
+   "widgets": {
+    "catalog_name": {
+     "currentValue": "dev_catalog",
+     "nuid": "74cb3499-b670-4a9b-a326-ad9c7a804698",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "dev_catalog",
+      "label": null,
+      "name": "catalog_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "dev_catalog",
+      "label": null,
+      "name": "catalog_name",
+      "options": {
+       "autoCreated": false,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "model_name": {
+     "currentValue": "quickstart_agent",
+     "nuid": "bfb488ca-a384-4fb5-a7ab-d05661a9d52d",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "quickstart_agent",
+      "label": null,
+      "name": "model_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "quickstart_agent",
+      "label": null,
+      "name": "model_name",
+      "options": {
+       "autoCreated": false,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "schema_name": {
+     "currentValue": "default",
+     "nuid": "1b42e31d-1ff1-4bf6-8a06-b55c595e45a9",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "default",
+      "label": null,
+      "name": "schema_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "default",
+      "label": null,
+      "name": "schema_name",
+      "options": {
+       "autoCreated": false,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    }
+   }
+  },
+  "language_info": {
+   "name": "python"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
+}
diff --git a/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/model_preprocessing.ipynb b/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/model_preprocessing.ipynb
new file mode 100644
index 0000000..95dd292
--- /dev/null
+++ b/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/model_preprocessing.ipynb
@@ -0,0 +1,222 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {},
+     "inputWidgets": {},
+     "nuid": "97addce4-2562-486a-bc4c-2b39a8dd9756",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "source": [
+    "## Generating inference data if it doesn't exist"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "c6242b0e-2772-410e-b7e9-b0bebe58d018",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "dbutils.widgets.text(\"catalog_name\", \"dev_catalog\")\n",
+    "dbutils.widgets.text(\"schema_name\", \"default\")\n",
+    "catalog_name = dbutils.widgets.get(\"catalog_name\")\n",
+    "schema_name = dbutils.widgets.get(\"schema_name\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "1e892951-2eef-40d1-b3f8-0a3be588a4d9",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "# Create the table with two columns: ID and review\n",
+    "spark.sql(f\"CREATE OR REPLACE TABLE {catalog_name}.{schema_name}.sentences (ID INT, review STRING)\")\n",
+    "\n",
+    "# Insert 200 values into the table with proper movie review text\n",
+    "movie_reviews = [\n",
+    "    \"A thrilling masterpiece with stunning visuals.\",\n",
+    "    \"An emotional rollercoaster that captivates from start to finish.\",\n",
+    "    \"A lackluster plot with underwhelming performances.\",\n",
+    "    \"A heartwarming tale that resonates with audiences of all ages.\",\n",
+    "    \"A groundbreaking film that redefines the genre.\",\n",
+    "    \"A predictable storyline with clichéd characters.\",\n",
+    "    \"A visually stunning film with a compelling narrative.\",\n",
+    "    \"An inspiring story of hope and resilience.\",\n",
+    "    \"A forgettable movie with a weak script.\",\n",
+    "    \"A charming and delightful film that entertains throughout.\"\n",
+    "]\n",
+    "\n",
+    "# Repeat the reviews to fill 200 entries\n",
+    "reviews_to_insert = (movie_reviews * (200 // len(movie_reviews) + 1))[:200]\n",
+    "\n",
+    "spark.sql(f\"INSERT INTO {catalog_name}.{schema_name}.sentences (ID, review) VALUES\" + \", \".join([f\"({i}, '{reviews_to_insert[i-1]}')\" for i in range(1, 201)]) + \";\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "34cc0494-a537-4e1d-933e-a123da2d70b2",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "display(spark.read.table(f\"{catalog_name}.{schema_name}.sentences\"))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {},
+     "inputWidgets": {},
+     "nuid": "a2ef7511-a2d9-4259-b890-fda7a99182e8",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "source": [
+    "## Generating labelled data if it doesn't exist"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "9d9f0618-fa45-4541-a0ed-b50cfe5257a6",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "# Create the labelled table with an additional sentiment column\n",
+    "spark.sql(f\"CREATE OR REPLACE TABLE {catalog_name}.{schema_name}.labelled_sentences AS SELECT ID,  review, ai_analyze_sentiment(review) AS sentiment FROM {catalog_name}.{schema_name}.sentences\")\n",
+    "\n",
+    "# Display the new table\n",
+    "display(spark.read.table(f\"{catalog_name}.{schema_name}.labelled_sentences\"))\n"
+   ]
+  }
+ ],
+ "metadata": {
+  "application/vnd.databricks.v1+notebook": {
+   "computePreferences": null,
+   "dashboards": [],
+   "environmentMetadata": {
+    "base_environment": "",
+    "environment_version": "3"
+   },
+   "inputWidgetPreferences": null,
+   "language": "python",
+   "notebookMetadata": {
+    "pythonIndentUnit": 2
+   },
+   "notebookName": "model_preprocessing",
+   "widgets": {
+    "catalog_name": {
+     "currentValue": "dev_catalog",
+     "nuid": "c27b5912-2ca9-46c0-ae51-7f7dcf693134",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "dev_catalog",
+      "label": null,
+      "name": "catalog_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "dev_catalog",
+      "label": null,
+      "name": "catalog_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "schema_name": {
+     "currentValue": "default",
+     "nuid": "25f2fe7e-e320-4793-a8cb-2fe969ec7d13",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "default",
+      "label": null,
+      "name": "schema_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "default",
+      "label": null,
+      "name": "schema_name",
+      "options": {
+       "autoCreated": false,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    }
+   }
+  },
+  "language_info": {
+   "name": "python"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
+}
diff --git a/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/quickstart_agent.py b/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/quickstart_agent.py
new file mode 100644
index 0000000..9fb782f
--- /dev/null
+++ b/contrib/templates/llmops-basic/template/{{.project_name}}/notebooks/quickstart_agent.py
@@ -0,0 +1,74 @@
+import json
+import uuid
+from databricks.sdk import WorkspaceClient
+from databricks_openai import UCFunctionToolkit, DatabricksFunctionClient
+from typing import Any, Optional, Iterator
+
+import mlflow
+from mlflow.pyfunc import ChatAgent
+from mlflow.types.agent import ChatAgentMessage, ChatAgentResponse, ChatContext
+
+# Get an OpenAI client configured to talk to Databricks model serving endpoints
+# We'll use this to query an LLM in our agent
+openai_client = WorkspaceClient().serving_endpoints.get_open_ai_client()
+
+# The snippet below tries to pick the first LLM API available in your Databricks workspace
+# from a set of candidates. You can override and simplify it
+# to just specify LLM_ENDPOINT_NAME.
+LLM_ENDPOINT_NAME = "databricks-claude-3-7-sonnet"
+
+# Enable automatic tracing of LLM calls
+mlflow.openai.autolog()
+
+
+@mlflow.trace
+def run_agent(content, system_prompt=None):
+    """
+    Send a user prompt to the LLM, and return a list of LLM response messages
+    The LLM is allowed to call the code interpreter tool if needed, to respond to the user
+    """
+    messages = []
+    if system_prompt:
+        messages.append({"role": "system", "content": system_prompt})
+    
+    messages.append({"role": "user", "content": content})
+    
+    response = openai_client.chat.completions.create(
+        model=LLM_ENDPOINT_NAME,
+        messages=messages,
+    )
+    msg = response.choices[0].message
+    return [msg.to_dict()]
+
+
+class QuickstartAgent(ChatAgent):
+    def __init__(self):
+        super().__init__()
+        self.system_prompt = (
+            "You are a sentiment analysis expert. Analyze text sentiment "
+            "and respond with exactly one word: positive, neutral, or negative."
+        )
+    
+    def predict(
+        self,
+        messages: list[ChatAgentMessage],
+        context: Optional[ChatContext] = None,
+        custom_inputs: Optional[dict[str, Any]] = None,
+    ) -> ChatAgentResponse:
+        message = messages[-1].content
+        user_prompt = f"Analyze this phrase: {message}"
+        raw_msgs = run_agent(
+            content=user_prompt,
+            system_prompt=self.system_prompt
+        )
+        out = []
+        for m in raw_msgs:
+            out.append(ChatAgentMessage(
+                id=uuid.uuid4().hex,
+                **m
+            ))
+        return ChatAgentResponse(messages=out)
+
+
+AGENT = QuickstartAgent()
+mlflow.models.set_model(AGENT)
diff --git a/contrib/templates/llmops-basic/template/{{.project_name}}/resources/model_artifacts.yml.tmpl b/contrib/templates/llmops-basic/template/{{.project_name}}/resources/model_artifacts.yml.tmpl
new file mode 100644
index 0000000..bcb7891
--- /dev/null
+++ b/contrib/templates/llmops-basic/template/{{.project_name}}/resources/model_artifacts.yml.tmpl
@@ -0,0 +1,21 @@
+resources:
+  registered_models:
+    sentiment_model:
+      name: ${var.model_name}
+      catalog_name: ${var.catalog_name}
+      schema_name: ${resources.schemas.sentiment_schema.name}
+      comment: Registered model in Unity Catalog for ${bundle.target} deployment target
+      grants:
+        - privileges:
+            - EXECUTE
+          principal: account users
+
+  experiments:
+    experiment:
+      name: ${var.experiment_name}
+
+  schemas:
+    sentiment_schema:
+      name: ${var.schema_name}
+      catalog_name: ${var.catalog_name}
+
diff --git a/contrib/templates/llmops-basic/template/{{.project_name}}/resources/model_build_evaluation.job.yml.tmpl b/contrib/templates/llmops-basic/template/{{.project_name}}/resources/model_build_evaluation.job.yml.tmpl
new file mode 100644
index 0000000..5633175
--- /dev/null
+++ b/contrib/templates/llmops-basic/template/{{.project_name}}/resources/model_build_evaluation.job.yml.tmpl
@@ -0,0 +1,32 @@
+resources:
+  jobs:
+    model_build_evaluation:
+      name: model_build_evaluation
+      queue:
+        enabled: true
+
+      parameters:
+        - name: "catalog_name"
+          default: "${var.catalog_name}"
+        - name: "schema_name"
+          default: "${resources.schemas.sentiment_schema.name}"    
+        - name: "model_name"
+          default: "${resources.registered_models.sentiment_model.name}"
+        - name: "experiment_name"
+          default: "${resources.experiments.experiment.name}"
+          
+      tasks:
+        - task_key: model_build
+          notebook_task:
+            notebook_path: "../notebooks/model_build.ipynb"
+              
+        - task_key: model_evaluation
+          depends_on:
+            - task_key: model_build
+          notebook_task:
+            notebook_path: "../notebooks/model_evaluation.ipynb"
+            base_parameters:
+              logged_run_id: "{{"{{"}}tasks.model_build.values.logged_run_id{{"}}"}}"         
+      tags:
+        Project: "${bundle.name}"
+
diff --git a/contrib/templates/llmops-basic/template/{{.project_name}}/resources/model_endpoint_deploy.job.yml.tmpl b/contrib/templates/llmops-basic/template/{{.project_name}}/resources/model_endpoint_deploy.job.yml.tmpl
new file mode 100644
index 0000000..a357d7a
--- /dev/null
+++ b/contrib/templates/llmops-basic/template/{{.project_name}}/resources/model_endpoint_deploy.job.yml.tmpl
@@ -0,0 +1,25 @@
+resources:
+  jobs:
+    model_deploy:
+      name: model_endpoint_deploy
+      queue:
+        enabled: true
+
+      parameters:
+        - name: "catalog_name"
+          default: "${var.catalog_name}"
+        - name: "schema_name"
+          default: "${resources.schemas.sentiment_schema.name}"    
+        - name: "model_name"
+          default: "${resources.registered_models.sentiment_model.name}"
+        - name: "experiment_name"
+          default: "${resources.experiments.experiment.name}"
+
+      tasks:
+        - task_key: model_endpoint_deploy
+          notebook_task:
+            notebook_path: "../notebooks/model_endpoint_deploy.ipynb"
+
+      tags:
+        Project: "${bundle.name}"
+
diff --git a/contrib/templates/llmops-basic/template/{{.project_name}}/resources/model_inference.job.yml.tmpl b/contrib/templates/llmops-basic/template/{{.project_name}}/resources/model_inference.job.yml.tmpl
new file mode 100644
index 0000000..5246476
--- /dev/null
+++ b/contrib/templates/llmops-basic/template/{{.project_name}}/resources/model_inference.job.yml.tmpl
@@ -0,0 +1,25 @@
+resources:
+  jobs:
+    model_inference:
+      name: model_inference
+      queue:
+        enabled: true
+
+      parameters:
+        - name: "catalog_name"
+          default: "${var.catalog_name}"
+        - name: "schema_name"
+          default: "${resources.schemas.sentiment_schema.name}"    
+        - name: "model_name"
+          default: "${resources.registered_models.sentiment_model.name}"
+        - name: "experiment_name"
+          default: "${resources.experiments.experiment.name}"
+
+      tasks:
+        - task_key: model_inference
+          notebook_task:
+            notebook_path: "../notebooks/model_inference.ipynb"
+
+      tags:
+        Project: "${bundle.name}"
+
diff --git a/contrib/templates/llmops-basic/template/{{.project_name}}/resources/model_preprocessing.job.yml.tmpl b/contrib/templates/llmops-basic/template/{{.project_name}}/resources/model_preprocessing.job.yml.tmpl
new file mode 100644
index 0000000..d359088
--- /dev/null
+++ b/contrib/templates/llmops-basic/template/{{.project_name}}/resources/model_preprocessing.job.yml.tmpl
@@ -0,0 +1,21 @@
+resources:
+  jobs:
+    model_preprocessing:
+      name: model_preprocessing
+      queue:
+        enabled: true
+
+      parameters:
+        - name: "catalog_name"
+          default: "${var.catalog_name}"
+        - name: "schema_name"
+          default: "${resources.schemas.sentiment_schema.name}"
+
+      tasks:
+        - task_key: model_preprocessing
+          notebook_task:
+            notebook_path: "../notebooks/model_preprocessing.ipynb"
+  
+      tags:
+        Project: "${bundle.name}"
+
diff --git a/llmops_basic/.gitignore b/llmops_basic/.gitignore
new file mode 100644
index 0000000..f3b15af
--- /dev/null
+++ b/llmops_basic/.gitignore
@@ -0,0 +1,22 @@
+.databricks/
+.bundle/
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+.Python
+venv/
+env/
+ENV/
+.ipynb_checkpoints/
+*.ipynb_checkpoints
+.vscode/
+.idea/
+*.swp
+*.swo
+*~
+.DS_Store
+Thumbs.db
+mlruns/
+*.log
+
diff --git a/llmops_basic/README.md b/llmops_basic/README.md
new file mode 100644
index 0000000..b0f7a71
--- /dev/null
+++ b/llmops_basic/README.md
@@ -0,0 +1,44 @@
+# llmops_basic
+
+End-to-end LLMOps project for sentiment analysis of reviews using Databricks.
+
+## Overview
+
+This example demonstrates a complete LLMOps pipeline for building, evaluating, and deploying a sentiment analysis model.
+
+**Pipeline stages:**
+- Data preparation
+- Model build and evaluation with MLflow
+- Model deployment to serving endpoint
+- Batch inference
+
+## Requirements
+
+- Databricks CLI (v0.218.0+)
+- Unity Catalog enabled
+- Required permissions on the catalog:
+  - `USE CATALOG` - to access the catalog
+  - `CREATE SCHEMA` - to create schemas
+
+## Quick Start
+
+**Deploy:**
+```bash
+databricks bundle deploy -t dev
+```
+
+**Run pipeline:**
+```bash
+databricks bundle run model_preprocessing -t dev
+databricks bundle run model_build_evaluation -t dev
+databricks bundle run model_endpoint_deploy -t dev
+databricks bundle run model_inference -t dev
+```
+
+## Configuration
+
+- **Dev Catalog**: `dev_catalog`
+- **Prod Catalog**: `prod_catalog`
+
+Edit `databricks.yml` to customize catalog names, schema name, model name, and experiment settings.
+
diff --git a/llmops_basic/databricks.yml b/llmops_basic/databricks.yml
new file mode 100644
index 0000000..a68bac3
--- /dev/null
+++ b/llmops_basic/databricks.yml
@@ -0,0 +1,43 @@
+# This is a Databricks Asset Bundle for llmops_basic.
+bundle:
+  name: "llmops_basic"
+
+variables:
+  catalog_name:
+    description: "Name of the UC catalog to use"
+    default: "default_catalog"
+  schema_name:
+    description: "Name of the UC schema to use"
+    default: "sentiment_agent_project"
+  model_name:
+    description: "Name of the UC model to use"
+    default: "sentiment_agent"
+  experiment_name:
+    description: "Name of experiment"
+    default: "/Users/${workspace.current_user.userName}/${bundle.target}_sentiment_agent"
+
+include:
+  - resources/*.yml
+
+targets:
+  dev:
+    mode: development
+    default: true
+    workspace:
+      host: https://your-workspace.azuredatabricks.net/
+    variables:
+      catalog_name: dev_catalog
+
+  prod:
+    mode: production
+    workspace:
+      host: https://your-workspace.azuredatabricks.net/
+      root_path: /Shared/.bundle/prod/${bundle.name}
+    variables:
+      catalog_name: prod_catalog
+    run_as:
+      # This runs as user@company.com in production. Alternatively,
+      # a service principal could be used here using service_principal_name
+      # (see Databricks documentation).
+      user_name: user@company.com
+    
\ No newline at end of file
diff --git a/llmops_basic/notebooks/model_build.ipynb b/llmops_basic/notebooks/model_build.ipynb
new file mode 100644
index 0000000..8ea1c35
--- /dev/null
+++ b/llmops_basic/notebooks/model_build.ipynb
@@ -0,0 +1,276 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "64ed089a-931a-4123-afdc-a2c38390774c",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "%pip install -U -qqqq mlflow databricks-openai databricks-agents\n",
+    "dbutils.library.restartPython()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "7f453049-63d9-46b2-8421-92f8c6a5442a",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "dbutils.widgets.text(\"catalog_name\", \"dev_catalog\")\n",
+    "dbutils.widgets.text(\"schema_name\", \"default\")\n",
+    "dbutils.widgets.text(\"model_name\", \"quickstart_agent\")\n",
+    "\n",
+    "\n",
+    "catalog_name = dbutils.widgets.get(\"catalog_name\")\n",
+    "schema_name = dbutils.widgets.get(\"schema_name\")\n",
+    "model_name = dbutils.widgets.get(\"model_name\")\n",
+    "dbutils.widgets.text(\"experiment_name\", f\"/Users/{dbutils.notebook.entry_point.getDbutils().notebook().getContext().userName().get()}/{model_name}_{catalog_name}\")\n",
+    "experiment_name = dbutils.widgets.get(\"experiment_name\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "114153c0-b167-451b-a2ac-6d4f48faeaec",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "import mlflow\n",
+    "\n",
+    "registered_model_name = f\"{catalog_name}.{schema_name}.{model_name}\"\n",
+    "\n",
+    "mlflow.set_experiment(experiment_name)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "154fbd28-5ad2-4a51-8ccf-fa850dffc5bd",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "from mlflow.models.resources import DatabricksFunction, DatabricksServingEndpoint\n",
+    "from pkg_resources import get_distribution\n",
+    "from quickstart_agent import LLM_ENDPOINT_NAME\n",
+    "import datetime\n",
+    "\n",
+    "mlflow.set_registry_uri(\"databricks-uc\")\n",
+    "resources = [\n",
+    "    DatabricksServingEndpoint(endpoint_name=LLM_ENDPOINT_NAME),\n",
+    "]\n",
+    "\n",
+    "timestamp = datetime.datetime.now().strftime(\"%Y%m%d_%H%M%S\")\n",
+    "\n",
+    "with mlflow.start_run(run_name=f\"sentiment_agent_{timestamp}\"):\n",
+    "    logged_model_info = mlflow.pyfunc.log_model(\n",
+    "        artifact_path=\"agent\",\n",
+    "        python_model=\"quickstart_agent.py\",\n",
+    "        extra_pip_requirements=[\n",
+    "            f\"databricks-connect=={get_distribution('databricks-connect').version}\"\n",
+    "        ],\n",
+    "        resources=resources\n",
+    "    )"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "ea348fb4-9a9f-4772-aaf1-65a646ca5739",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "dbutils.jobs.taskValues.set(key = \"logged_run_id\", value = logged_model_info.run_id)"
+   ]
+  }
+ ],
+ "metadata": {
+  "application/vnd.databricks.v1+notebook": {
+   "computePreferences": {
+    "hardware": {
+     "accelerator": null,
+     "gpuPoolId": null,
+     "memory": null
+    }
+   },
+   "dashboards": [],
+   "environmentMetadata": {
+    "base_environment": "",
+    "environment_version": "3"
+   },
+   "inputWidgetPreferences": null,
+   "language": "python",
+   "notebookMetadata": {
+    "pythonIndentUnit": 2
+   },
+   "notebookName": "model_build",
+   "widgets": {
+    "catalog_name": {
+     "currentValue": "dev_catalog",
+     "nuid": "fd954ea2-6905-4526-b686-38b43c5b6d97",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "dev_catalog",
+      "label": null,
+      "name": "catalog_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "dev_catalog",
+      "label": null,
+      "name": "catalog_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "experiment_name": {
+     "currentValue": "",
+     "nuid": "818e024c-c654-4a9c-9f36-7e595b9a9e5f",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "tee",
+      "label": null,
+      "name": "experiment_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "tee",
+      "label": null,
+      "name": "experiment_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "model_name": {
+     "currentValue": "quickstart_agent",
+     "nuid": "93dd77ea-7b35-410c-af71-589d260ae712",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "quickstart_agent",
+      "label": null,
+      "name": "model_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "quickstart_agent",
+      "label": null,
+      "name": "model_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "schema_name": {
+     "currentValue": "default",
+     "nuid": "88bd61ef-aa5c-42a2-ad26-f9079c1afc5f",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "default",
+      "label": null,
+      "name": "schema_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "default",
+      "label": null,
+      "name": "schema_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    }
+   }
+  },
+  "language_info": {
+   "name": "python"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
+}
diff --git a/llmops_basic/notebooks/model_endpoint_deploy.ipynb b/llmops_basic/notebooks/model_endpoint_deploy.ipynb
new file mode 100644
index 0000000..2f18515
--- /dev/null
+++ b/llmops_basic/notebooks/model_endpoint_deploy.ipynb
@@ -0,0 +1,118 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {},
+     "inputWidgets": {},
+     "nuid": "28525745-4fb6-4667-92f1-1e0e3fa68da1",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "%pip install -U -qqqq mlflow databricks-openai databricks-agents\n",
+    "dbutils.library.restartPython()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {},
+     "inputWidgets": {},
+     "nuid": "b2519e1e-ef7e-439c-ab55-5c17e77c8b25",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "dbutils.widgets.text(\"catalog_name\", \"dev_catalog\")\n",
+    "dbutils.widgets.text(\"schema_name\", \"default\")\n",
+    "dbutils.widgets.text(\"model_name\", \"quickstart_agent\")\n",
+    "\n",
+    "catalog_name = dbutils.widgets.get(\"catalog_name\")\n",
+    "schema_name = dbutils.widgets.get(\"schema_name\")\n",
+    "model_name = dbutils.widgets.get(\"model_name\")\n",
+    "dbutils.widgets.text(\"experiment_name\", f\"/Users/{dbutils.notebook.entry_point.getDbutils().notebook().getContext().userName().get()}/{model_name}_{catalog_name}\")\n",
+    "experiment_name = dbutils.widgets.get(\"experiment_name\")\n",
+    "\n",
+    "import mlflow\n",
+    "\n",
+    "mlflow.set_experiment(experiment_name)\n",
+    "\n",
+    "registered_model_name = f\"{catalog_name}.{schema_name}.{model_name}\""
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {},
+     "inputWidgets": {},
+     "nuid": "b2a5f820-740b-4e9d-8496-4b54c3350156",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "#https://docs.databricks.com/aws/en/generative-ai/agent-framework/deploy-agent#deploy-an-agent-using-deploy\n",
+    "\n",
+    "from mlflow import MlflowClient\n",
+    "from databricks.agents import get_deployments, delete_deployment\n",
+    "from databricks import agents\n",
+    "\n",
+    "client = MlflowClient()\n",
+    "\n",
+    "registered_model_version = client.get_model_version_by_alias(registered_model_name, \"Champion\")\n",
+    "\n",
+    "deployments = get_deployments(model_name=registered_model_name)\n",
+    "\n",
+    "latest_version = registered_model_version.version\n",
+    "\n",
+    "for deployment in deployments:\n",
+    "    if deployment.model_name == registered_model_name:\n",
+    "        print(f\"Deleting deployment: model={deployment.model_name}, version={deployment.model_version}\")\n",
+    "        delete_deployment(model_name=deployment.model_name, model_version=deployment.model_version)\n",
+    "\n",
+    "\n",
+    "deployment_info = agents.deploy(\n",
+    "    model_name=registered_model_name,\n",
+    "    model_version=registered_model_version.version,\n",
+    ")"
+   ]
+  }
+ ],
+ "metadata": {
+  "application/vnd.databricks.v1+notebook": {
+   "computePreferences": null,
+   "dashboards": [],
+   "environmentMetadata": {
+    "base_environment": "",
+    "environment_version": "3"
+   },
+   "inputWidgetPreferences": null,
+   "language": "python",
+   "notebookMetadata": {
+    "pythonIndentUnit": 2
+   },
+   "notebookName": "model_endpoint_deploy",
+   "widgets": {}
+  },
+  "language_info": {
+   "name": "python"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
+}
diff --git a/llmops_basic/notebooks/model_evaluation.ipynb b/llmops_basic/notebooks/model_evaluation.ipynb
new file mode 100644
index 0000000..5d08f41
--- /dev/null
+++ b/llmops_basic/notebooks/model_evaluation.ipynb
@@ -0,0 +1,396 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "be63921b-4820-4872-a708-bf5722e0f8e7",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "%pip install -U -qqqq mlflow databricks-openai databricks-agents threadpoolctl==3.1.0\n",
+    "dbutils.library.restartPython()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "8af65798-0acc-4263-bc35-9212653d3987",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "dbutils.widgets.text(\"catalog_name\", \"dev_catalog\")\n",
+    "dbutils.widgets.text(\"schema_name\", \"default\")\n",
+    "dbutils.widgets.text(\"model_name\", \"quickstart_agent\")\n",
+    "dbutils.widgets.text(\"logged_run_id\", \"None\")\n",
+    "\n",
+    "catalog_name = dbutils.widgets.get(\"catalog_name\")\n",
+    "schema_name = dbutils.widgets.get(\"schema_name\")\n",
+    "model_name = dbutils.widgets.get(\"model_name\")\n",
+    "logged_run_id = dbutils.widgets.get(\"logged_run_id\")\n",
+    "\n",
+    "dbutils.widgets.text(\"experiment_name\", f\"/Users/{dbutils.notebook.entry_point.getDbutils().notebook().getContext().userName().get()}/{model_name}_{catalog_name}\")\n",
+    "experiment_name = dbutils.widgets.get(\"experiment_name\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "fec9759b-b50d-4611-be17-d3b51a3041a6",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "import mlflow\n",
+    "import mlflow.genai.datasets\n",
+    "import time\n",
+    "from databricks.connect import DatabricksSession\n",
+    "\n",
+    "registered_model_name = f\"{catalog_name}.{schema_name}.{model_name}\"\n",
+    "model_uri = f\"runs:/{logged_run_id}/agent\"\n",
+    "\n",
+    "mlflow.set_experiment(experiment_name)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "f0a33257-a9da-4f65-8c8a-657612d6d634",
+     "showTitle": false,
+     "tableResultSettingsMap": {
+      "0": {
+       "dataGridStateBlob": "{\"version\":1,\"tableState\":{\"columnPinning\":{\"left\":[\"#row_number#\"],\"right\":[]},\"columnSizing\":{},\"columnVisibility\":{}},\"settings\":{\"columns\":{}},\"syncTimestamp\":1756760635746}",
+       "filterBlob": null,
+       "queryPlanFiltersBlob": null,
+       "tableResultIndex": 0
+      }
+     },
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "query = f\"SELECT * FROM {catalog_name}.{schema_name}.labelled_sentences\"\n",
+    "df = spark.sql(query).toPandas()\n",
+    "display(df)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "0c997521-a041-46ed-8b37-b0b9ddf67df3",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "import mlflow\n",
+    "from mlflow.genai.scorers import Guidelines, Correctness, RelevanceToQuery\n",
+    "import datetime\n",
+    "\n",
+    "eval_data = []\n",
+    "for _, row in df.iterrows():\n",
+    "    eval_item = {\n",
+    "        \"inputs\": {\n",
+    "            \"content\": row[\"review\"]  # This matches the function parameter name\n",
+    "        },\n",
+    "        \"expectations\": {\n",
+    "            \"expected_response\": str(row[\"sentiment\"])  # Adjust column name\n",
+    "        }\n",
+    "    }\n",
+    "    eval_data.append(eval_item)\n",
+    "\n",
+    "mlflow_eval_dataset = None\n",
+    "    \n",
+    "try:\n",
+    "    # Try to get existing dataset\n",
+    "    mlflow_eval_dataset = mlflow.genai.get_dataset(f\"{catalog_name}.{schema_name}.mlflow_eval_dataset\")\n",
+    "    print(f\"✓ Using existing dataset: {catalog_name}.{schema_name}.mlflow_eval_dataset\")\n",
+    "    \n",
+    "except Exception as get_error:\n",
+    "    print(f\"Dataset not found, creating dataset\")\n",
+    "    \n",
+    "    try:\n",
+    "        # Create new dataset\n",
+    "        mlflow_eval_dataset = mlflow.genai.create_dataset(f\"{catalog_name}.{schema_name}.mlflow_eval_dataset\")\n",
+    "        mlflow_eval_dataset.merge_records(eval_data)\n",
+    "        dataset_created = True\n",
+    "        print(f\"✓ Created new dataset: {catalog_name}.{schema_name}.mlflow_eval_dataset\")\n",
+    "        \n",
+    "    except Exception as create_error:\n",
+    "        print(f\"⚠ Could not create dataset {create_error}\")\n",
+    "\n",
+    "if (mlflow_eval_dataset is not None):       \n",
+    "        \n",
+    "    guidelines = {\n",
+    "        \"sentiment_accuracy\": \"Response must correctly identify sentiment\",\n",
+    "        \"clarity\": [\"Response must be clear and concise\"]\n",
+    "    }\n",
+    "\n",
+    "    agent = mlflow.pyfunc.load_model(model_uri)\n",
+    "\n",
+    "    def predict_function(content):\n",
+    "        try:\n",
+    "            messages = [\n",
+    "                {\n",
+    "                    \"role\": \"user\",\n",
+    "                    \"content\": content\n",
+    "                }\n",
+    "            ]\n",
+    "            prediction = agent.predict({\"messages\": messages})\n",
+    "            return prediction\n",
+    "        except Exception as e:\n",
+    "            print(f\"Error in predict_fn: {e}\")\n",
+    "            return {\"response\": \"PREDICTION_ERROR\"}\n",
+    "        \n",
+    "    timestamp = datetime.datetime.now().strftime(\"%Y%m%d_%H%M%S\")\n",
+    "    with mlflow.start_run(run_id=logged_run_id) as run:\n",
+    "        with mlflow.start_run(run_name=f\"sentiment_agent_evaluation_{timestamp}\", nested=True) as eval_run:\n",
+    "            results = mlflow.genai.evaluate(\n",
+    "                data=mlflow_eval_dataset,\n",
+    "                predict_fn=predict_function,\n",
+    "                scorers=[\n",
+    "                    Correctness(),\n",
+    "                    Guidelines(name=\"sentiment_accuracy\", guidelines=guidelines[\"sentiment_accuracy\"]),\n",
+    "                    Guidelines(name=\"clarity\", guidelines=guidelines[\"clarity\"]),\n",
+    "                ],\n",
+    "            )\n",
+    "\n",
+    "    # Print results\n",
+    "    print(\"Evaluation Results:\")\n",
+    "    print(results.metrics)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {},
+     "inputWidgets": {},
+     "nuid": "281e07cf-0163-453a-9256-c508b37cd073",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "from mlflow import MlflowClient\n",
+    "client = MlflowClient()\n",
+    "\n",
+    "registered_model = mlflow.register_model(model_uri, name=registered_model_name)\n",
+    "\n",
+    "if results.metrics['sentiment_accuracy/mean'] > 0.9:\n",
+    "  print(\"Transitioning to champion\")\n",
+    "  client.set_registered_model_alias(registered_model_name, \"Champion\",registered_model.version)"
+   ]
+  }
+ ],
+ "metadata": {
+  "application/vnd.databricks.v1+notebook": {
+   "computePreferences": {
+    "hardware": {
+     "accelerator": null,
+     "gpuPoolId": null,
+     "memory": null
+    }
+   },
+   "dashboards": [],
+   "environmentMetadata": {
+    "base_environment": "",
+    "environment_version": "3"
+   },
+   "inputWidgetPreferences": null,
+   "language": "python",
+   "notebookMetadata": {
+    "pythonIndentUnit": 2
+   },
+   "notebookName": "model_evaluation",
+   "widgets": {
+    "catalog_name": {
+     "currentValue": "dev_catalog",
+     "nuid": "3a964089-b210-4dc6-aa79-c7c287530d31",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "dev_catalog",
+      "label": null,
+      "name": "catalog_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "dev_catalog",
+      "label": null,
+      "name": "catalog_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "experiment_name": {
+     "currentValue": "/Users/user@company.com/quickstart_agent_dev_catalog",
+     "nuid": "d243520b-4f07-4843-b294-0e3b3489d85b",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "/Users/user@company.com/quickstart_agent_dev_catalog",
+      "label": null,
+      "name": "experiment_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "/Users/user@company.com/quickstart_agent_dev_catalog",
+      "label": null,
+      "name": "experiment_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "logged_run_id": {
+     "currentValue": "None",
+     "nuid": "970893b9-068c-4fff-af42-e9b346618f94",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "None",
+      "label": null,
+      "name": "logged_run_id",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "None",
+      "label": null,
+      "name": "logged_run_id",
+      "options": {
+       "autoCreated": false,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "model_name": {
+     "currentValue": "quickstart_agent",
+     "nuid": "06926edb-a96f-4f02-a702-f7c542e5ef3c",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "quickstart_agent",
+      "label": null,
+      "name": "model_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "quickstart_agent",
+      "label": null,
+      "name": "model_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "schema_name": {
+     "currentValue": "default",
+     "nuid": "febc8306-c467-4694-9bfe-6467bb3558a2",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "default",
+      "label": null,
+      "name": "schema_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "default",
+      "label": null,
+      "name": "schema_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    }
+   }
+  },
+  "language_info": {
+   "name": "python"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
+}
diff --git a/llmops_basic/notebooks/model_inference.ipynb b/llmops_basic/notebooks/model_inference.ipynb
new file mode 100644
index 0000000..e0c4535
--- /dev/null
+++ b/llmops_basic/notebooks/model_inference.ipynb
@@ -0,0 +1,243 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "be63921b-4820-4872-a708-bf5722e0f8e7",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "%pip install -U -qqqq mlflow databricks-openai databricks-agents\n",
+    "dbutils.library.restartPython()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "138d2537-42eb-4fae-82b8-e11021a8a85b",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "dbutils.widgets.text(\"catalog_name\", \"dev_catalog\")\n",
+    "dbutils.widgets.text(\"schema_name\", \"default\")\n",
+    "dbutils.widgets.text(\"model_name\", \"quickstart_agent\")\n",
+    "catalog_name = dbutils.widgets.get(\"catalog_name\")\n",
+    "schema_name = dbutils.widgets.get(\"schema_name\")\n",
+    "model_name = dbutils.widgets.get(\"model_name\")\n",
+    "\n",
+    "dbutils.widgets.text(\"experiment_name\", f\"/Users/{dbutils.notebook.entry_point.getDbutils().notebook().getContext().userName().get()}/{model_name}_{catalog_name}\")\n",
+    "experiment_name = dbutils.widgets.get(\"experiment_name\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "fec9759b-b50d-4611-be17-d3b51a3041a6",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "import mlflow\n",
+    "import mlflow.genai.datasets\n",
+    "import time\n",
+    "from databricks.connect import DatabricksSession\n",
+    "\n",
+    "mlflow.set_experiment(experiment_name)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "40bbaf35-a61d-439f-9bf0-b160a96223ab",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "query = f\"SELECT * FROM {catalog_name}.{schema_name}.sentences\"\n",
+    "df = spark.sql(query).toPandas()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "0c997521-a041-46ed-8b37-b0b9ddf67df3",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "\n",
+    "model_uri = f\"models:/{catalog_name}.{schema_name}.{model_name}@champion\"\n",
+    "agent = mlflow.pyfunc.load_model(model_uri)\n",
+    "\n",
+    "def predict_sentiment(review):\n",
+    "    messages = [{\n",
+    "        \"role\": \"user\",\n",
+    "        \"content\": f\"{review}\"\n",
+    "    }]\n",
+    "    return agent.predict({\"messages\": messages})\n",
+    "\n",
+    "df['predictions'] = df['review'].apply(predict_sentiment)\n",
+    "display(df)\n"
+   ]
+  }
+ ],
+ "metadata": {
+  "application/vnd.databricks.v1+notebook": {
+   "computePreferences": {
+    "hardware": {
+     "accelerator": null,
+     "gpuPoolId": null,
+     "memory": null
+    }
+   },
+   "dashboards": [],
+   "environmentMetadata": {
+    "base_environment": "",
+    "environment_version": "3"
+   },
+   "inputWidgetPreferences": null,
+   "language": "python",
+   "notebookMetadata": {
+    "pythonIndentUnit": 2
+   },
+   "notebookName": "model_inference",
+   "widgets": {
+    "catalog_name": {
+     "currentValue": "dev_catalog",
+     "nuid": "74cb3499-b670-4a9b-a326-ad9c7a804698",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "dev_catalog",
+      "label": null,
+      "name": "catalog_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "dev_catalog",
+      "label": null,
+      "name": "catalog_name",
+      "options": {
+       "autoCreated": false,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "model_name": {
+     "currentValue": "quickstart_agent",
+     "nuid": "bfb488ca-a384-4fb5-a7ab-d05661a9d52d",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "quickstart_agent",
+      "label": null,
+      "name": "model_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "quickstart_agent",
+      "label": null,
+      "name": "model_name",
+      "options": {
+       "autoCreated": false,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "schema_name": {
+     "currentValue": "default",
+     "nuid": "1b42e31d-1ff1-4bf6-8a06-b55c595e45a9",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "default",
+      "label": null,
+      "name": "schema_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "default",
+      "label": null,
+      "name": "schema_name",
+      "options": {
+       "autoCreated": false,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    }
+   }
+  },
+  "language_info": {
+   "name": "python"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
+}
diff --git a/llmops_basic/notebooks/model_preprocessing.ipynb b/llmops_basic/notebooks/model_preprocessing.ipynb
new file mode 100644
index 0000000..95dd292
--- /dev/null
+++ b/llmops_basic/notebooks/model_preprocessing.ipynb
@@ -0,0 +1,222 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {},
+     "inputWidgets": {},
+     "nuid": "97addce4-2562-486a-bc4c-2b39a8dd9756",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "source": [
+    "## Generating inference data if it doesn't exist"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "c6242b0e-2772-410e-b7e9-b0bebe58d018",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "dbutils.widgets.text(\"catalog_name\", \"dev_catalog\")\n",
+    "dbutils.widgets.text(\"schema_name\", \"default\")\n",
+    "catalog_name = dbutils.widgets.get(\"catalog_name\")\n",
+    "schema_name = dbutils.widgets.get(\"schema_name\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "1e892951-2eef-40d1-b3f8-0a3be588a4d9",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "# Create the table with two columns: ID and review\n",
+    "spark.sql(f\"CREATE OR REPLACE TABLE {catalog_name}.{schema_name}.sentences (ID INT, review STRING)\")\n",
+    "\n",
+    "# Insert 200 values into the table with proper movie review text\n",
+    "movie_reviews = [\n",
+    "    \"A thrilling masterpiece with stunning visuals.\",\n",
+    "    \"An emotional rollercoaster that captivates from start to finish.\",\n",
+    "    \"A lackluster plot with underwhelming performances.\",\n",
+    "    \"A heartwarming tale that resonates with audiences of all ages.\",\n",
+    "    \"A groundbreaking film that redefines the genre.\",\n",
+    "    \"A predictable storyline with clichéd characters.\",\n",
+    "    \"A visually stunning film with a compelling narrative.\",\n",
+    "    \"An inspiring story of hope and resilience.\",\n",
+    "    \"A forgettable movie with a weak script.\",\n",
+    "    \"A charming and delightful film that entertains throughout.\"\n",
+    "]\n",
+    "\n",
+    "# Repeat the reviews to fill 200 entries\n",
+    "reviews_to_insert = (movie_reviews * (200 // len(movie_reviews) + 1))[:200]\n",
+    "\n",
+    "spark.sql(f\"INSERT INTO {catalog_name}.{schema_name}.sentences (ID, review) VALUES\" + \", \".join([f\"({i}, '{reviews_to_insert[i-1]}')\" for i in range(1, 201)]) + \";\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "34cc0494-a537-4e1d-933e-a123da2d70b2",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "display(spark.read.table(f\"{catalog_name}.{schema_name}.sentences\"))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {},
+     "inputWidgets": {},
+     "nuid": "a2ef7511-a2d9-4259-b890-fda7a99182e8",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "source": [
+    "## Generating labelled data if it doesn't exist"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 0,
+   "metadata": {
+    "application/vnd.databricks.v1+cell": {
+     "cellMetadata": {
+      "byteLimit": 2048000,
+      "rowLimit": 10000
+     },
+     "inputWidgets": {},
+     "nuid": "9d9f0618-fa45-4541-a0ed-b50cfe5257a6",
+     "showTitle": false,
+     "tableResultSettingsMap": {},
+     "title": ""
+    }
+   },
+   "outputs": [],
+   "source": [
+    "# Create the labelled table with an additional sentiment column\n",
+    "spark.sql(f\"CREATE OR REPLACE TABLE {catalog_name}.{schema_name}.labelled_sentences AS SELECT ID,  review, ai_analyze_sentiment(review) AS sentiment FROM {catalog_name}.{schema_name}.sentences\")\n",
+    "\n",
+    "# Display the new table\n",
+    "display(spark.read.table(f\"{catalog_name}.{schema_name}.labelled_sentences\"))\n"
+   ]
+  }
+ ],
+ "metadata": {
+  "application/vnd.databricks.v1+notebook": {
+   "computePreferences": null,
+   "dashboards": [],
+   "environmentMetadata": {
+    "base_environment": "",
+    "environment_version": "3"
+   },
+   "inputWidgetPreferences": null,
+   "language": "python",
+   "notebookMetadata": {
+    "pythonIndentUnit": 2
+   },
+   "notebookName": "model_preprocessing",
+   "widgets": {
+    "catalog_name": {
+     "currentValue": "dev_catalog",
+     "nuid": "c27b5912-2ca9-46c0-ae51-7f7dcf693134",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "dev_catalog",
+      "label": null,
+      "name": "catalog_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "dev_catalog",
+      "label": null,
+      "name": "catalog_name",
+      "options": {
+       "autoCreated": null,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    },
+    "schema_name": {
+     "currentValue": "default",
+     "nuid": "25f2fe7e-e320-4793-a8cb-2fe969ec7d13",
+     "typedWidgetInfo": {
+      "autoCreated": false,
+      "defaultValue": "default",
+      "label": null,
+      "name": "schema_name",
+      "options": {
+       "validationRegex": null,
+       "widgetDisplayType": "Text"
+      },
+      "parameterDataType": "String"
+     },
+     "widgetInfo": {
+      "defaultValue": "default",
+      "label": null,
+      "name": "schema_name",
+      "options": {
+       "autoCreated": false,
+       "validationRegex": null,
+       "widgetType": "text"
+      },
+      "widgetType": "text"
+     }
+    }
+   }
+  },
+  "language_info": {
+   "name": "python"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
+}
diff --git a/llmops_basic/notebooks/quickstart_agent.py b/llmops_basic/notebooks/quickstart_agent.py
new file mode 100644
index 0000000..9fb782f
--- /dev/null
+++ b/llmops_basic/notebooks/quickstart_agent.py
@@ -0,0 +1,74 @@
+import json
+import uuid
+from databricks.sdk import WorkspaceClient
+from databricks_openai import UCFunctionToolkit, DatabricksFunctionClient
+from typing import Any, Optional, Iterator
+
+import mlflow
+from mlflow.pyfunc import ChatAgent
+from mlflow.types.agent import ChatAgentMessage, ChatAgentResponse, ChatContext
+
+# Get an OpenAI client configured to talk to Databricks model serving endpoints
+# We'll use this to query an LLM in our agent
+openai_client = WorkspaceClient().serving_endpoints.get_open_ai_client()
+
+# The snippet below tries to pick the first LLM API available in your Databricks workspace
+# from a set of candidates. You can override and simplify it
+# to just specify LLM_ENDPOINT_NAME.
+LLM_ENDPOINT_NAME = "databricks-claude-3-7-sonnet"
+
+# Enable automatic tracing of LLM calls
+mlflow.openai.autolog()
+
+
+@mlflow.trace
+def run_agent(content, system_prompt=None):
+    """
+    Send a user prompt to the LLM, and return a list of LLM response messages
+    The LLM is allowed to call the code interpreter tool if needed, to respond to the user
+    """
+    messages = []
+    if system_prompt:
+        messages.append({"role": "system", "content": system_prompt})
+    
+    messages.append({"role": "user", "content": content})
+    
+    response = openai_client.chat.completions.create(
+        model=LLM_ENDPOINT_NAME,
+        messages=messages,
+    )
+    msg = response.choices[0].message
+    return [msg.to_dict()]
+
+
+class QuickstartAgent(ChatAgent):
+    def __init__(self):
+        super().__init__()
+        self.system_prompt = (
+            "You are a sentiment analysis expert. Analyze text sentiment "
+            "and respond with exactly one word: positive, neutral, or negative."
+        )
+    
+    def predict(
+        self,
+        messages: list[ChatAgentMessage],
+        context: Optional[ChatContext] = None,
+        custom_inputs: Optional[dict[str, Any]] = None,
+    ) -> ChatAgentResponse:
+        message = messages[-1].content
+        user_prompt = f"Analyze this phrase: {message}"
+        raw_msgs = run_agent(
+            content=user_prompt,
+            system_prompt=self.system_prompt
+        )
+        out = []
+        for m in raw_msgs:
+            out.append(ChatAgentMessage(
+                id=uuid.uuid4().hex,
+                **m
+            ))
+        return ChatAgentResponse(messages=out)
+
+
+AGENT = QuickstartAgent()
+mlflow.models.set_model(AGENT)
diff --git a/llmops_basic/resources/model_artifacts.yml b/llmops_basic/resources/model_artifacts.yml
new file mode 100644
index 0000000..bcb7891
--- /dev/null
+++ b/llmops_basic/resources/model_artifacts.yml
@@ -0,0 +1,21 @@
+resources:
+  registered_models:
+    sentiment_model:
+      name: ${var.model_name}
+      catalog_name: ${var.catalog_name}
+      schema_name: ${resources.schemas.sentiment_schema.name}
+      comment: Registered model in Unity Catalog for ${bundle.target} deployment target
+      grants:
+        - privileges:
+            - EXECUTE
+          principal: account users
+
+  experiments:
+    experiment:
+      name: ${var.experiment_name}
+
+  schemas:
+    sentiment_schema:
+      name: ${var.schema_name}
+      catalog_name: ${var.catalog_name}
+
diff --git a/llmops_basic/resources/model_build_evaluation.job.yml b/llmops_basic/resources/model_build_evaluation.job.yml
new file mode 100644
index 0000000..856d808
--- /dev/null
+++ b/llmops_basic/resources/model_build_evaluation.job.yml
@@ -0,0 +1,32 @@
+resources:
+  jobs:
+    model_build_evaluation:
+      name: model_build_evaluation
+      queue:
+        enabled: true
+
+      parameters:
+        - name: "catalog_name"
+          default: "${var.catalog_name}"
+        - name: "schema_name"
+          default: "${resources.schemas.sentiment_schema.name}"    
+        - name: "model_name"
+          default: "${resources.registered_models.sentiment_model.name}"
+        - name: "experiment_name"
+          default: "${resources.experiments.experiment.name}"
+          
+      tasks:
+        - task_key: model_build
+          notebook_task:
+            notebook_path: "../notebooks/model_build.ipynb"
+              
+        - task_key: model_evaluation
+          depends_on:
+            - task_key: model_build
+          notebook_task:
+            notebook_path: "../notebooks/model_evaluation.ipynb"
+            base_parameters:
+              logged_run_id: "{{tasks.model_build.values.logged_run_id}}"         
+      tags:
+        Project: "${bundle.name}"
+
diff --git a/llmops_basic/resources/model_endpoint_deploy.job.yml b/llmops_basic/resources/model_endpoint_deploy.job.yml
new file mode 100644
index 0000000..a357d7a
--- /dev/null
+++ b/llmops_basic/resources/model_endpoint_deploy.job.yml
@@ -0,0 +1,25 @@
+resources:
+  jobs:
+    model_deploy:
+      name: model_endpoint_deploy
+      queue:
+        enabled: true
+
+      parameters:
+        - name: "catalog_name"
+          default: "${var.catalog_name}"
+        - name: "schema_name"
+          default: "${resources.schemas.sentiment_schema.name}"    
+        - name: "model_name"
+          default: "${resources.registered_models.sentiment_model.name}"
+        - name: "experiment_name"
+          default: "${resources.experiments.experiment.name}"
+
+      tasks:
+        - task_key: model_endpoint_deploy
+          notebook_task:
+            notebook_path: "../notebooks/model_endpoint_deploy.ipynb"
+
+      tags:
+        Project: "${bundle.name}"
+
diff --git a/llmops_basic/resources/model_inference.job.yml b/llmops_basic/resources/model_inference.job.yml
new file mode 100644
index 0000000..5246476
--- /dev/null
+++ b/llmops_basic/resources/model_inference.job.yml
@@ -0,0 +1,25 @@
+resources:
+  jobs:
+    model_inference:
+      name: model_inference
+      queue:
+        enabled: true
+
+      parameters:
+        - name: "catalog_name"
+          default: "${var.catalog_name}"
+        - name: "schema_name"
+          default: "${resources.schemas.sentiment_schema.name}"    
+        - name: "model_name"
+          default: "${resources.registered_models.sentiment_model.name}"
+        - name: "experiment_name"
+          default: "${resources.experiments.experiment.name}"
+
+      tasks:
+        - task_key: model_inference
+          notebook_task:
+            notebook_path: "../notebooks/model_inference.ipynb"
+
+      tags:
+        Project: "${bundle.name}"
+
diff --git a/llmops_basic/resources/model_preprocessing.job.yml b/llmops_basic/resources/model_preprocessing.job.yml
new file mode 100644
index 0000000..d359088
--- /dev/null
+++ b/llmops_basic/resources/model_preprocessing.job.yml
@@ -0,0 +1,21 @@
+resources:
+  jobs:
+    model_preprocessing:
+      name: model_preprocessing
+      queue:
+        enabled: true
+
+      parameters:
+        - name: "catalog_name"
+          default: "${var.catalog_name}"
+        - name: "schema_name"
+          default: "${resources.schemas.sentiment_schema.name}"
+
+      tasks:
+        - task_key: model_preprocessing
+          notebook_task:
+            notebook_path: "../notebooks/model_preprocessing.ipynb"
+  
+      tags:
+        Project: "${bundle.name}"
+