I build Python automation, AI-assisted tools, and production-ready data pipelines. I like systems that are fast, reliable, and easy to debug.
⁛ Selected projects
- Panopticon CVE Map
open-source— desktop app for vulnerability analysis with CISA KEV + local AI agent. dev.to. - Murmur
open-source— Zero-Knowledge secret scanner for Git hooks (SHA-256, no raw secrets). - Aero
open-source— Windows File Explorer opacity control with atomic config writes. - Market Analytics Bot
open-source— real-time arbitrage analytics with MCP server and Streamlit dashboard. - Sanity
private— high-frequency trading bot with WebSocket + REST fallback, 0 HTTP 429 on 56M+ items. Architecture documented in README. - GooEye
private— Goofish sniper bot with multi-user isolation, mtop API interception, 0 CAPTCHA on 44k+ items.
⁛ What I enjoy building
- Async systems that survive edge cases (WebSocket + REST fallback, retry with backoff)
- AI agents with structured tools (MCP servers, function calling, evals)
- Data pipelines that don't lose data (atomic writes, deduplication, cleanup)
- Desktop apps with local-first storage (SQLite, no cloud dependency)
- Security tools that respect privacy (Zero-Knowledge, no telemetry)
- Systems with contracts — clear inputs, predictable outputs, no magic
⁛ Currently working on
- 🚀 Panopticon CVE Map — CISA KEV integration, MD export, versioned chat (article published on dev.to)
- 🔒 Murmur — Zero-Knowledge secret scanner for Git hooks (pre-commit security layer)
- 📚 Learning: LangGraph, RAG architecture, eval frameworks (Langfuse/Ragas), ClickHouse