I'm an AI Product Manager who builds side projects and infrastructure to deeply understand the systems I manage. I focus on cutting through the noise to ship practical, scalable AI products.

Most of my hands-on work here revolves around prototyping, self-hosting models, and exploring how emerging tech actually behaves in practice—rather than just how it looks in a pitch deck.

Applied AI & Knowledge Graphs: Building AutoK, a personal knowledge-graph project to explore data relationships, memory, and practical LLM utility.

Frameworks & API Generation: loom – An AI-first, design-first API framework in Go (forked from Goa). It generates strict OpenAPI 3.1 contracts and handles multi-protocol transport (HTTP, gRPC, JSON-RPC) to provide trustworthy specs for agents and client generators.

Agentic Runtimes & Tooling: loom-mcp – A runtime and code generator leveraging the Model Context Protocol (MCP) to define agents, toolsets, and registries, handling everything from planning to durable execution.

Agent Skills & Workflow Design: skills – A growing collection of reusable agent skills for Claude Code and similar coding agents. It captures focused workflows for testing, refactoring, frontend cleanup, Go upgrade reviews, TypeDB/TypeQL, and SurrealDB to make agent behavior more specialized, reliable, and useful in real codebases.

Knowledge Graphs & Data Modeling: go-typeql – A Go ORM for TypeDB 3.x. It maps graph schemas to Go structs, wrapping the Rust driver via CGo to deliver type-safe CRUD, query building, and schema migrations. (This powers the data architecture for my personal knowledge-graph project, Auto-K).

Token-Efficient Diagnostics: remedy – A Go library for structured business errors. It packages HTTP status, machine-readable codes, and inline recovery guidance directly into the error, helping agentic systems self-correct without wasting tokens on follow-up retries.

Infrastructure & Self-Hosting: Running a custom homelab for deploying, testing, and understanding local AI models, hardware, and network architecture.

Senior Product Manager with 15+ years of experience specializing in AI and software development, with a track record at Google, Meta, and LinkedIn.

Writing & Thinking: I document my projects, product frameworks, and strategic thinking at luckymethod.com.

Building: I'm an active maker outside of software, splitting my time between woodworking, 3D printing, and CNC machining.

🌱 Whether you're building early-stage AI startups, experimenting with local LLMs, or tackling complex product strategy, I'm always open to connect and discuss how I can be helpful.