I build high-stakes AI products from scratch. While I operate at the frontiers of technology with CERN and Carnegie Mellon, my true obsession is the product lifecycle: identifying pain points, architecting scalable solutions, and shipping software that "people want."
| Project | Role | Description | Tech Stack |
|---|---|---|---|
| Postwand | Co-Founder | AI SaaS automating social media for businesses. Managed full lifecycle from legal LLC incorporation to official Meta/Twitter API vetting. | React (Vite), Python/Flask, Celery, Redis, OpenAI |
| GetBank | Full-Stack | High-complexity Solana lending protocol. Engineered interest rate models and liquidation engines with mid-six-figure budget scope. | Rust (Anchor), Java/Spring, Pyth Oracles, Angular |
- Core Languages: Python (Expert), Rust, TypeScript, C++, Java, SQL.
- AI & Agentic Systems: LangGraph, LangChain, RAG, Multi-Agent Orchestration, Langfuse, Vector DBs.
- Full-Stack & Systems: Next.js 16, FastAPI, React (Zustand), Spark (Distributed Computing), CUDA (H100/T4).
- Execution & Infra: Docker, Redis, Celery, GitHub Actions, Linux, UI/UX (Figma).
- CERN (LHCb): Engineering a RAG-enhanced agentic framework to automate the restoration of legacy physics pipelines. Leveraging MCP (Model Context Protocol) with Cursor to synthesize decades of fragmented documentation.
- CMU Research: Building high-performance training pipelines to investigate Scaling Laws and Dropout Universality. Focused on hardware-level optimization using CUDA.
- Accenture (Ex-Data & AI): Developed specialized LLMs for Tier-1 global banking compliance, replacing legacy systems with enterprise-grade RAG architectures.
"The best way to predict the future is to build it."
I am a product-first engineer. I follow the Y Combinator school of thought: iterate fast, talk to users, and prioritize shipping velocity over perfect abstraction. My background in Data Science and Computer Engineering allows me to bridge the gap between complex mathematical research and production-ready code.
