Regional Director @ Plan A Technologies · AI Toolmaker · 15+ Years Shipping Software

I spent 15 years building software the hard way. Now I build the tools so you can ship AI agents in days, not months.

Chat UI, runtime, tools, memory, RAG, and observability behind one API instead of eight libraries. Swap OpenAI → Claude, React → terminal, in-memory → a vector DB — without rewriting your code.

npm install @agentskit/core @agentskit/react @agentskit/runtime

Core · React UI · Terminal UI · Runtime · Tools · Skills · Memory · RAG · Observability — all composable, all independent.

The rest of the stack:

• Enterprise AI Playbook — hard-won patterns for teams putting AI to work: rollout, governance, agent architecture, and the mistakes that stall most projects before production.
• AKOS — AgentsKit OS — orchestrate, govern, and audit every agent your team runs. Configured in YAML, a GUI, or code, deployed on your own infra.
• Registry — the shadcn for agents. Ready-to-use agents, tools, and skills you install and own — no lock-in, just building blocks that work.

AgentsKit to build it. The Registry to skip the boilerplate. AKOS to run it. The Playbook to do it right.

The resource I wish I had when I started with AI. Code examples, comparison tables, architecture diagrams. Curated learning paths from beginner to production-ready.

From idea to revenue, solo. Ideation, architecture, development, go-to-market, operations — every hat a builder needs, automated. Works with Claude Code, Cursor, Copilot, and 30+ tools.

• Regional Director at Plan A Technologies — leading 180+ engineers across LATAM
• 500+ technical interviews conducted — watched the industry shift to AI firsthand
• 15+ years shipping software: Military → Freelancer → Government → Senior Dev → Tech Lead → Architect → Director
• Founder of NoxCode — building AI-powered products with a small team

Every week I break down one AI engineering topic with the depth of someone who's shipped it in production.

DS&A, Big O, system design and AI system design (RAG, agents, evals, cost, safety), prompt engineering, ML fundamentals, behavioral, and negotiation. Written from both sides of the table — 500+ interviews, 180+ engineers led. Available in English and Portuguese.