About
BuildingagenticAIsystemsandfine-tuningLLMsforproductionuse.I help companies move from AI experimentation to production-grade systems. I build systems that reason, call tools, persist state, and ship to production.
Fine-Tuning & Model Training
Training and adapting open-source LLMs for domain-specific use. Full pipeline: data collection, distillation, QLoRA training, evaluation, iteration. Making models that don't hallucinate.
- Python
- Unsloth
- HuggingFace
- QLoRA
- Claude API
Agentic Systems
Designing AI agents that reason, call tools, and take action. Multi-user architectures with auth, state persistence, voice input, and REST APIs. Built with LangGraph, deployed to production.
- LangGraph
- FastAPI
- PostgreSQL
- Auth0
- Railway
AI Product Delivery
Full-stack AI products: Python backends, Docker containers, cloud GPU orchestration, payment integration. Architected, built, and deployed solo.
- Docker
- RunPod
- Blender
- Stripe
- Three.js
- Unreal Engine
Currently developing VR healthcare training simulations for enterprise clients. Previously delivered for Philips, Tommy Hilfiger, KitKat, and Porsche Design.