AI-native engineer based in Chicago, IL. I build the tools that let people work with AI agents: desktop apps, MCP infrastructure, and developer productivity software, from first prototype to shipped product.
| Project | Description | Stars |
|---|---|---|
| Korus | Agentic Development Environment (Electron + Vue) | 🆕 |
| Skwad | AI coding agents, working together (Swift) | 🆕 |
| Witsy | Desktop AI assistant & universal MCP client (Electron + Vue) | ⭐ 2k+ |
| multi-llm-ts | Unified TypeScript library for LLM provider APIs | ⭐ 80+ |
- Nestor — API Hub for AI Agents
- rag-youtube — Build a RAG from YouTube channel videos
- youtube-tldw — YouTube TL;DW — video summarization with LLMs
- Trax — macOS music management app (Flutter)
- foto — Image browser/viewer (Flutter)
- easy-chapters — macOS app to edit video chapters
- vue-hero-transition — Hero transition component for Vue
- taglib_ffi — TagLib FFI bindings for Dart/Flutter
- powerpoint-custom-table-style-generator — Custom styles for PowerPoint tables
I build with agents every day, and Korus and Skwad are where I figure out how to do it well.
- Orchestrating teams of agents — running several coding agents in parallel on the same codebase, each with its own role and context
- Agent-ready workflows — worktrees, tests, and reviews that let agents move fast without breaking things
- Tools and protocols agents rely on — MCP servers and clients, multi-provider LLM abstractions
- Language-agnostic shipping — TypeScript, Swift, Python, Flutter: I pick what fits the product and let agents handle the syntax