added openai support and put into docker container

.env.example

@@ -1,4 +1,5 @@
 # Environment Configuration Template
+# Copy to .env and fill in secrets. .env is gitignored.
 
 # Database Configuration
 # SQLite (default):
@@ -10,10 +11,21 @@ DATABASE_URL=sqlite:///./asktemoc.db
 # Enable database query logging
 DB_ECHO=false
 
-# ChromaDB Configuration
+# ChromaDB Configuration (required for /api/chat RAG)
+# Use /app/chroma_db when running in Docker.
 CHROMA_PERSIST_DIRECTORY=./app/chroma_db
 CHROMA_COLLECTION_NAME=asktemoc_collection
 
+# OpenAI (optional; if set, /api/chat uses OpenAI instead of Ollama)
+OPENAI_API_KEY=
+OPENAI_MODEL=gpt-4o-mini
+OPENAI_EMBEDDING_MODEL=text-embedding-3-small
+
+# Ollama (used when OPENAI_API_KEY is not set)
+# OLLAMA_BASE_URL=http://localhost:11434
+# OLLAMA_MODEL=llama3.1:8b
+# OLLAMA_EMBEDDING_MODEL=nomic-embed-text
+
 # FastAPI Configuration
 ENVIRONMENT=development
 DEBUG=true

.gitignore

@@ -1,4 +1,5 @@
 /venv
+.env
 
 __pycache__
 

Dockerfile

@@ -0,0 +1,29 @@
+FROM python:3.11-slim
+
+ENV PYTHONUNBUFFERED=1 \
+    PYTHONDONTWRITEBYTECODE=1 \
+    PIP_NO_CACHE_DIR=1 \
+    PIP_DISABLE_PIP_VERSION_CHECK=1 \
+    PIP_ROOT_USER_ACTION=ignore
+
+WORKDIR /app
+
+# System deps (if you add more heavy libs later, extend this)
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    build-essential \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install a minimal, container-focused dependency set to avoid resolver issues
+# from the much larger local development requirements.
+COPY requirements-docker.txt /app/requirements-docker.txt
+RUN pip install --upgrade pip && pip install -r /app/requirements-docker.txt
+
+# Copy application code
+COPY . /app
+
+# Expose FastAPI default port
+EXPOSE 8000
+
+# Uvicorn entrypoint
+CMD ["uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "8000"]
+

app/services/llm_service.py

@@ -1,17 +1,19 @@
 from langchain_ollama import ChatOllama
 import random
+
+
 class LLMService:
     def __init__(self):
         self.llm = ChatOllama(model="llama3.1:8b", temperature=0.4)
 
     def call(self):
-        # note that this is just a place holder, this class is only for calling the LLM 
+        # note that this is just a place holder, this class is only for calling the LLM
         countries = ["Indonesia", "Germany", "China", "France", "Japan", "Brazil"]
-        content = self.llm.invoke(f"What is the capital of {random.choices(countries)}?").content
+        content = self.llm.invoke(f"What is the capital of {random.choice(countries)}?").content
         return content
 
     async def a_call(self):
         # Just know that this is an async call to the LLM hence the 'a_' prefix
         countries = ["Indonesia", "Germany", "China", "France", "Japan", "Brazil"]
-        content = await self.llm.ainvoke(f"What is the capital of {random.choices(countries)}?")
+        content = await self.llm.ainvoke(f"What is the capital of {random.choice(countries)}?")
         return content.content

app/services/rag_chain_service.py

@@ -1,15 +1,31 @@
+import os
+
 from langchain_community.llms import Ollama
-from langchain_core.runnables import RunnablePassthrough, RunnableParallel
 from langchain_core.output_parsers import StrOutputParser
+from langchain_core.runnables import RunnableParallel, RunnablePassthrough
 
 from app.services.prompt_service import rag_prompt_template
 from app.services.retriever_service import retriever_service
-import os
+
+
+def _get_llm():
+    """Use OpenAI if OPENAI_API_KEY is set, otherwise Ollama."""
+    if os.getenv("OPENAI_API_KEY"):
+        from langchain_openai import ChatOpenAI
+        return ChatOpenAI(
+            model=os.getenv("OPENAI_MODEL", "gpt-4o-mini"),
+            temperature=0.4,
+        )
+    return Ollama(
+        model=os.getenv("OLLAMA_MODEL", "llama3.1:8b"),
+        base_url=os.getenv("OLLAMA_BASE_URL", "http://localhost:11434"),
+    )
+
 
 class RagChainService:
     def __init__(self):
         self.retriever = retriever_service.get_retriever()
-        self.llm = Ollama(model=os.getenv("OLLAMA_MODEL", "llama3.1:8b"), base_url=os.getenv("OLLAMA_BASE_URL", "http://localhost:11434"))
+        self.llm = _get_llm()
 
     def get_chain(self):
         def format_docs(docs):

app/services/retriever_service.py

@@ -1,13 +1,25 @@
+import os
+
 import chromadb
 from langchain_chroma import Chroma
-from langchain_community.embeddings import OllamaEmbeddings
-import os
+
+
+def _get_embeddings():
+    """Use OpenAI embeddings if OPENAI_API_KEY is set, otherwise Ollama."""
+    if os.getenv("OPENAI_API_KEY"):
+        from langchain_openai import OpenAIEmbeddings
+        return OpenAIEmbeddings(model=os.getenv("OPENAI_EMBEDDING_MODEL", "text-embedding-3-small"))
+    from langchain_community.embeddings import OllamaEmbeddings
+    return OllamaEmbeddings(model=os.getenv("OLLAMA_EMBEDDING_MODEL", "nomic-embed-text"))
+
 
 class RetrieverService:
     def __init__(self, collection_name="asktemoc_collection"):
-        self.client = chromadb.PersistentClient(path=os.getenv("CHROMA_PERSIST_DIRECTORY"))
+        self.client = chromadb.PersistentClient(
+            path=os.getenv("CHROMA_PERSIST_DIRECTORY", "./app/chroma_db")
+        )
         self.collection_name = collection_name
-        self.embeddings = OllamaEmbeddings(model=os.getenv("OLLAMA_EMBEDDING_MODEL", "nomic-embed-text"))
+        self.embeddings = _get_embeddings()
 
     def get_retriever(self):
         vector_store = Chroma(

requirements-docker.txt

@@ -0,0 +1,18 @@
+fastapi
+uvicorn[standard]
+SQLAlchemy
+pydantic
+pydantic-settings
+python-dotenv
+requests
+
+# RAG / LLM stack
+langchain>=0.3.0
+langchain-community
+langchain-chroma
+langchain-ollama
+langchain-openai
+chromadb
+openai>=1.0.0
+crawl4ai
+