AI Engineer building agent systems, local-first AI tools, and LLM-powered products.
I like turning AI experiments into software people can actually use.
I'm currently focused on AI agents, local AI, context engineering, and human–AI interfaces — especially systems that go beyond a simple LLM chat wrapper.
- Agent runtimes, tool use, and multi-agent workflows
- Local speech / multimodal AI
- LLM gateways and OpenAI-compatible infrastructure
- Full-stack AI products from prototype to desktop / web application
🎙️ VoxCue
A local-first voice input assistant for ChatGPT, built with Electron, React, TypeScript, and sherpa-onnx.
It combines local speech recognition, VAD, voiceprint verification, conversation switching, and desktop integration into a practical AI interface.
⚙️ ollama_mock
An Ollama-compatible LLM gateway written in Python, with streaming responses, multi-provider routing, load balancing, and rate limiting across providers such as OpenRouter and Groq.
A lightweight Python desktop utility for synchronizing clipboard content between computers through a shared directory.
AI Agents · LLM Infrastructure · Local AI · RAG · Context Engineering · Speech AI · Python · TypeScript · Java
Build useful things with AI.