I build open-source developer tools and production systems that are easier to understand, operate, and own.
I work across Go, Python, Swift, and TypeScript, usually where product thinking meets systems engineering: agent-native tools, local-first software, modular backends, and practical infrastructure.
My goal is simple: turn complicated workflows into software people can inspect, trust, and keep running.
| Project | What it does | Built with |
|---|---|---|
| LeadShoot | Evidence-first local-business research. Qualifies leads without disguising unknowns as facts; ships as a CLI, REST API, MCP server, and live map. | Python · SQLite · FastAPI · MCP |
| Production SaaS Starter | A production-oriented B2B SaaS foundation built as a modular monolith with clear boundaries and portable infrastructure. | Go · Next.js · PostgreSQL |
| Nanowave | Generates, compiles, and runs SwiftUI apps from natural-language instructions through Claude Code. | Go · SwiftUI · Claude Code |
| iOSJumpstart | A SwiftUI starter with authentication, payments, backend, and notifications already integrated. | Swift 6 · Supabase · RevenueCat |
| iOS Dev Agent | Reusable skills, rules, and integrations for AI-assisted iOS development across major coding agents. | Python · MCP · Agent tooling |
evidence over confident guesses systems over disconnected features
ownership over unnecessary lock-in clear boundaries over accidental complexity
tools that work over demos that impress
I am especially interested in:
- Agent-native software — structured tools that agents can use reliably, with deterministic boundaries around probabilistic reasoning.
- Local-first and open systems — portable data, inspectable behavior, and infrastructure the operator controls.
- Production architecture — modular monoliths, stable interfaces, and systems designed around operational reality.
- Developer experience — strong defaults, small feedback loops, and documentation that gets people to a working result quickly.
- Making LeadShoot a trustworthy research layer for anyone selling to local businesses.
- Exploring better interfaces between AI agents and real development tools.
- Publishing reusable foundations for teams that want to ship without surrendering control of their stack.
If you are building an ambitious developer tool, agent workflow, or production system, I would like to hear about it.