Skip to content
View AbhayJuloori's full-sized avatar

Highlights

  • Pro

Block or report AbhayJuloori

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
AbhayJuloori/README.md

Abhay Juloori

I build production-style ML and analytics systems that are meant to be used, not just trained. Most of my work sits at the intersection of credit risk, survival analysis, anomaly detection, forecasting, and decision-support tooling, with an emphasis on taking projects from modeling through APIs and usable front ends.

I am especially interested in applied ML and data work where modeling, software, and product thinking meet.


What you'll find here

  • End-to-end ML projects grounded in real decision contexts
  • Data systems with APIs, dashboards, evaluation workflows, and documentation
  • Projects that balance statistical rigor, implementation detail, and usability

Selected projects

Project What it demonstrates
LoanSurv Survival analysis for loan default timing on 2.2M Lending Club loans using Cox PH + Random Survival Forest, served through FastAPI with a React dashboard. [live demo]
Credit Recourse Engine Actionable recourse for denied applicants using XGBoost, conformal prediction, and counterfactual explanations.
Freight KPI Tracker Freight analytics pipeline with synthetic generation seeded from FAF5, Snowflake loading, lane-level anomaly detection, and dashboarding.
Controlled Document Retrieval Compliance-focused retrieval system with role and jurisdiction filters, document lifecycle controls, and audit-ready citations.
codex-claude-bridge Local orchestration bridge where Claude handles planning and review while Codex executes implementation-heavy work.
opsmon Data reliability monitoring for multi-source time series with anomaly, drift, missingness, and integrity checks.

How I like to build

  • Start from a concrete decision or workflow, not just a benchmark
  • Make uncertainty, evaluation, and failure modes visible
  • Ship models with an interface: CLI, API, dashboard, or workflow
  • Document tradeoffs clearly enough that another engineer can pick the work up

Stack

ML/Data     Python · scikit-learn · XGBoost · lifelines · scikit-survival · PyTorch
Data        pandas · SQL · Spark · Snowflake · Azure Synapse · Cosmos DB · pyarrow
APIs        FastAPI · Pydantic · Docker
Frontend    React · Vite · Plotly/Dash · Recharts · Tailwind CSS

Pinned Loading

  1. codex-claude-bridge codex-claude-bridge Public

    Run Claude Code with Codex as the worker — no OpenAI API key needed, just ChatGPT Plus. Claude plans & reviews, Codex executes. Local proxy bridge + hybrid orchestration runtime.

    TypeScript

  2. credit-recourse-engine credit-recourse-engine Public

    Credit-risk recourse engine using XGBoost, conformal prediction, and counterfactual explanations for denied applicants

    Python

  3. freight-kpi-tracker freight-kpi-tracker Public

    Freight analytics pipeline with FAF5-seeded synthetic shipments, Snowflake loading, lane-level anomaly detection, and dashboard reporting

    HTML

  4. loansurv loansurv Public

    Survival analysis for loan default timing using Cox PH and Random Survival Forest, served through FastAPI with a React dashboard

    JavaScript

  5. opsmon opsmon Public

    Data reliability monitoring for multi-source time series with drift, missingness, anomaly, and integrity checks

    Python

  6. controlled-document-retrieval-compliance-tracking-system controlled-document-retrieval-compliance-tracking-system Public

    Compliance document retrieval system with lifecycle controls, jurisdiction filters, and audit-ready citations

    Python