🛡️ ClaimTrackr

AI-Powered Health Insurance Claim Evaluation

ClaimTrackr is an LLM-powered RAG system for evaluating health insurance claims against policy documents, exclusions, and supporting records. It combines semantic retrieval, FastAPI, Pinecone, and Docker to deliver a reproducible, deployment-ready workflow.

---

What it does

ClaimTrackr automates claim evaluation by retrieving relevant policy context and using an LLM to generate a verdict with supporting reasoning.

---

Tech Stack

Component	Technology
Language	Python 3.11
Backend	FastAPI, Uvicorn
Frontend	React, Vite, Tailwind CSS
Package Manager	uv
Vector Database	Pinecone
Embeddings	SentenceTransformers
Containerization	Docker, Docker Compose

---

High-Level Flow

1. Ingest policy documents and claim-related files.
2. Embed document text using sentence-transformers/all-mpnet-base-v2.
3. Store and retrieve context from Pinecone.
4. Evaluate the claim with an LLM.
5. Return a verdict, summary, and rejection reasons if applicable.

---

Project Structure

ClaimTrackr/
├── main.py
├── src/
├── frontend/
├── documents/
├── bills/
├── pyproject.toml
├── uv.lock
├── Dockerfile
├── docker-compose.yml
├── .env.example
└── .dockerignore

---

How to run

Option 1: Run with Docker image

Use this if you just want to start the app quickly.

git clone https://github.com/yourusername/claimtrackr.git
cd claimtrackr
cp .env.example .env
docker compose up -d

Open:

http://localhost:8000

Option 2: Run locally for development

Use this if you want to edit the code and test changes.

uv sync
cd frontend
npm install
npm run dev

Then run the backend in another terminal:

python main.py

After making changes

docker compose up -d --build

Stop the app

docker compose down

---

Docker / Deployment Highlights

ClaimTrackr uses a multi-stage Dockerfile so dependencies are installed in a builder stage and only the final runtime files are copied into the production image.

• Smaller runtime image
• Faster deployment
• Cleaner production container
• Reduced image content size from 919MB to 455MB in this setup

---

Author

Shivam Modi

Software Developer