GraphLens (MappingMoney)

GraphLens is a CS6242 (Data and Visual Analytics, Georgia Tech, Spring 2026) course project focused on exploring financial relationships in the CMS Open Payments ecosystem using graph analytics and interactive visualizations.

At A Glance

• Interactive analytics for CMS Open Payments data
• Temporal graph exploration across years and products
• Centrality, community, anomaly, and concentration analysis
• 3D network exploration and dashboard storytelling

Course Project Info

• Course: CS6242 - Data and Visual Analytics
• Term: Spring 2026
• Institution: Georgia Institute of Technology
• Team: 11

Team

• Aaryesh Deshpande
• Birendra Kumar Saye
• Gabriel Appiah
• Prabhjot Singh Ahluwalia
• Sanjana Suresh

Project Overview

GraphLens studies payment flows between manufacturers and physicians across years, products, and network structures using publicly available CMS Open Payments data.

The project combines:

• Temporal graph construction
• Community and centrality analysis
• Anomaly detection
• Dashboard-driven visual storytelling
• 3D network exploration

Tech Stack

• Frontend: Svelte, Vite, D3, Deck.gl, Plotly, Three.js
• Backend API: Python, FastAPI, Pydantic, Polars, Uvicorn
• Graph/Analytics: igraph, NetworkX, NumPy, SciPy, scikit-learn, UMAP, PyArrow
• Infra: Docker, GitHub Actions

Data Source

• CMS Open Payments Data: https://openpaymentsdata.cms.gov/
• CMS Open Payments Program: https://www.cms.gov/openpayments

Quick Start

1. Install Python dependencies from pyproject.toml.
2. Run the API (for example with Uvicorn against api.main:app).
3. In frontend/, install Node dependencies and run npm run dev.

Local Deployment (Short)

1. Backend API:

   Install Python deps, then run: uvicorn api.main:app --reload --port 8000

2. Frontend:

• cd frontend
• npm install
• npm run dev

GitHub Pages (Short)

Website is hosted on GitHub Pages at https://aaryesh-ad.github.io/graphlens/.

The deployment process is automated with GitHub Actions, which builds the frontend and publishes it to Pages on pushes to main/master when relevant files change.

• Frontend deploy is automated with GitHub Actions via .github/workflows/deploy-frontend-pages.yml.
• The workflow runs on pushes to main/master when files in frontend/** (or the workflow file) change.

Build uses:

• VITE_BASE_PATH=/<repo-name>/ for Pages routing
• VITE_API_BASE_URL from repository secrets

Output from frontend/dist is published to GitHub Pages.

The backend is hosted on Render and is not part of the Pages deployment (separate infrastructure). The frontend calls this API for data.

License

This project is released under the MIT License. See LICENSE for details.

Acknowledgements

• CMS Open Payments Program for providing the data and ecosystem to explore financial relationships in healthcare.
• The Georgia Tech CS6242 course and instructor (Dr. Max Roozbahani), TAs for guiding the project and providing resources on data analytics and visualization techniques.
• Open-source libraries and tools that made the development of GraphLens possible, including Svelte, D3, FastAPI, igraph, and many others.