Full-Stack Developer & ML Software Engineer based in San Francisco, CA. I like building things from scratch, breaking them, and making them better.
Most recently, I led R&D on a network intrusion detection system at a healthcare cybersecurity startup, where I built ensemble ML pipelines (XGBoost, Random Forest, LightGBM, Isolation Forest, Autoencoders) processing millions of network flows daily. Before that, I got my Master's in Business Analytics & AI from UT Dallas.
I'm a hackathon regular and I ship fast. AkashGuard, my autonomous self-healing cloud deployment agent, won 2nd place at the Open Agents Hackathon by Overclock Labs. I also built Squiker (squiker.com), a full-stack ad-tech platform, solo from the ground up.
ML & Data: Python, Scikit-learn, XGBoost, LightGBM, TensorFlow, PyTorch, SHAP, Pandas, NumPy
Backend: Node.js, FastAPI, Flask, PostgreSQL, MongoDB, Redis
Frontend: React, Next.js, TypeScript, Tailwind CSS
Cloud & Infra: AWS, Azure, Docker, Kubernetes, CI/CD, GitHub Actions
🥈 AkashGuard - Autonomous self-healing cloud deployment agent using Llama 3.3 70B. Monitors, diagnoses, and redeploys failed services on Akash Network. 2nd place, Open Agents Hackathon - Feb 2026
📊 DealGraph - AI-powered due diligence copilot for investors. Multi-agent pipelines on AWS Bedrock with Neo4j knowledge graph fact-checking. AWS x Anthropic x Datadog GenAI Hackathon - Feb 2026
📸 Specto - Snapchat Spectacles lens that auto-captures sports highlights with a rolling 30s buffer and AI-powered moment analysis. Gemini 3 SuperHack - Jan 2026
🎙️ Cactus Voice - Fully on-device voice-to-action app using Whisper + FunctionGemma for sub-millisecond intent routing. Cactus x Google DeepMind Hackathon - Feb 2026
⚖️ LegalStrategyCouncil - Multi-agent legal strategy system. Multi-agent collaboration using MongoDB Atlas as the coordination backbone. MongoDB Hackathon