Contract Review & Risk Analysis System (CAUD / CUAD)
ML Engineer / AI Engineer
Quick Links
On this page
AI-powered contract analysis and risk scoring system with live Streamlit dashboard + Cloud Run FastAPI API, batch processing, CI/CD, and monitoring.
Visuals
System architecture overview
Click image to view full-size
Top clause performance metrics
Click image to view full-size
Model calibration quality assessment
Click image to view full-size
ROI and cost-benefit analysis
Click image to view full-size
Problem
Manual contract review is slow and inconsistent. Legal teams need scalable clause detection, risk scoring, and portfolio-level insights with a production-ready workflow.
Solution
Implemented an end-to-end contract analysis pipeline supporting TXT/PDF ingestion (PyMuPDF), clause segmentation, TF-IDF feature extraction, ML risk classification with calibrated confidence, and portfolio-level analytics. Built a FastAPI backend with real-time inference + batch endpoints, and a Streamlit UI with rich visualizations (risk distribution, red-flag triage, portfolio insights). Containerized with Docker, deployed API on Google Cloud Run, and implemented CI/CD via GitHub Actions with Prometheus metrics + Grafana monitoring.
Key Results
- Live Streamlit dashboard + deployed Cloud Run API
- API response time: <500ms average
- Batch processing for multi-contract analysis
- Portfolio analytics + red-flag triage
- CI/CD via GitHub Actions
- Observability: Prometheus metrics + Grafana dashboards