Frontline (AI Medic)
We built a system that watches live video and listens to patients to figure out injuries instantly.
- The Goal: Automate patient triage using multi-modal AI (Vision + Audio).
- The Tech: React, Node.js, Gemini Vision API, Presage.
RAG AI Agent
I created a chatbot for Headstarter that uses Retrieval Augmented Generation (RAG) to give accurate responses rather than just hallucinating.
- The Goal: Improve customer support accuracy by grounding AI answers in real documentation.
- The Tech: Next.js, OpenAI, Pinecone (Vector DB).
High-Volume Data Parser
An efficient algorithm written in C to ingest large CSV datasets with over a million records.
- The Goal: optimize data retrieval speeds by building a custom in-memory hash index.
- The Tech: C, Pointers, Manual Memory Management.
Current Project: Real-Time Sentiment Pipeline I'm building a distributed system that tracks news sentiment in real-time.
- The Goal: Move away from "Monoliths" and build a proper microservices architecture.
- The Tech:
- Ingestor: A Python script scraping headlines.
- Queue: Pushing data into Redis (so I don't lose data if the server crashes).
- Worker: A separate service picking up jobs and saving them to PostgreSQL.
- Ops: Orchestrating it all with Docker Compose.
