AI Lab
Where AI meets production engineering
I treat LLMs and computer vision as engineering problems: real interfaces, real constraints, real trade-offs — not isolated experiments in a notebook.
LLMs & Agents
Multi-agent orchestration with CrewAI, RAG pipelines, and fine-tuning LLMs for domain-specific applications — most recently at Sensilog, building AI features for real client products.
Computer Vision
Real-time object detection and gesture-based interaction, built on YOLO, OpenCV, and MediaPipe — from academic research to interactive HCI experiments.
Applied, not academic
Every project here ties back to a concrete interface or workflow. The goal is always a usable system, not a standalone notebook.
AI & Computer Vision projects
Full case studies — problem, architecture, and trade-offs — for each project below.
