I’m Priyanshu Sah, an AI Engineer focused on turning AI research and models into products people can actually use.
Currently at Liferythm Healthcare, I work on an AI powered healthcare suite spanning clinical documentation, patient triage, AI receptionists, follow-up automation, React Native applications, FastAPI microservices, and Bluetooth enabled medical device integrations. My work spans the full product lifecycle from data preparation and evaluation to backend development, model integration, deployment, and production AI systems for community healthcare.
My interests include LLM applications, agentic systems, model serving, inference optimization, and building reliable software around AI models. I enjoy building AI systems end-to-end from research and experimentation to production-ready applications used in real-world workflows
- Agentic Systems | LangGraph | RAG | MCP
- Medical LLMs | Transformers | LangChain | Hugging Face
- Fine-Tuning | Evaluation Pipelines | Unsloth
- Multimodal AI + VLMs | Model Serving | Inference Optimization
- React Native | React | FastAPI | Flask
- REST APIs | WebSockets | Node.js
- System Architecture | Nginx | Cloudflare
- Docker | GitHub Actions | AWS | GCP
- CI/CD | Elasticsearch | Production Deployments
- Project Maintainer of TradingViewPlus, contributing features, architectural improvements, and adapter-based support for TradingView-powered platforms used by 300+ traders.
- Contributed 20+ pull requests, resolved 30+ issues, shipped major releases, and continue maintaining the project alongside the community.
| Project | Description | Tech Stack | 📄 Case Study |
|---|---|---|---|
| 🩺 Healthcare AI Platform (Liferythm) | End-to-end healthcare AI platform integrating React Native, FastAPI microservices, LangGraph workflows, Medical LLMs, and Bluetooth-enabled medical device integrations for community healthcare. | React Native · FastAPI · LangGraph · Medical LLMs | Case Study → |
| 🎙️ Enterprise Conversational AI Platform (CREO) | Enterprise Speech-to-Speech AI platform with configurable Voice AI assistants, retrieval-augmented generation, and reusable deployment architecture. | FastAPI · LangGraph · LangChain | Case Study → |
| 🎮 PokéWorld | Research sandbox exploring Dreamer-style world models, representation learning, latent dynamics, and model-based reinforcement learning. | PyTorch · Model-Based RL | Case Study → |
| 📈 TradingViewPlus | Open-source browser extension maintained for 300+ traders, featuring architectural improvements, adapter-based integrations, and cross-browser support. | TypeScript · WebExtensions API | Case Study → |
| 🤖 NavDP (Research Sandbox) | Research project extending Navigation Diffusion Policy with custom omni-directional robot support, simulator improvements, and sim-to-real experimentation. | Python · PyTorch · Robotics | Case Study → |
| 🧠 Enerzal | Enterprise Graph-RAG assistant automating IT and HR workflows through document understanding, tool calling, and secure authentication. | Flask · LangChain · Neo4j | Case Study → |
| 🌍 Eco Chain | Full-stack Web3 application for decentralized product provenance, ownership tracking, and blockchain-based supply chain transparency. | Solidity · Web3.js · React | Case Study → |
