At wobot.ai, I built a system that analyzes camera scenes and recommends best-fit agents from a suite of 120+ video AI agents. It also powers scene-type analytics and surfaces use cases the suite doesn't yet cover.
I built an agentic chatbot for a drag-and-drop video-analytics pipeline builder, backed by an automated test suite for incident response and bug fixes. It lets customer success teams configure pipelines directly.
At wobot.ai, I created an API that runs about 5,000 configuration validations concurrently in under a second. I also decoupled Docker dependency layers, cutting CI/CD build times by 73% and image sizes by about 1.2GB.
My open-source projects include Scibot, a provider-agnostic multi-agent RAG tool for scientific research, and Omnipost, a microservices social media orchestrator. I work with Python backend systems, agentic AI, and tools including Django DRF, FastAPI, PostgreSQL, MongoDB, and Redis.

