I help engineering teams get AI from demo to production. 15+ years of production software experience, former CTO, FDA-cleared medical device software, still shipping daily. Most AI demos never survive contact with production. My work focuses on what closes that gap: architecture, evals, observability, and the engineering judgment that makes agentic systems reliable.

Working patterns from real projects, not toy demos:

• agent-skills: the skills library I use daily with coding agents, installable as Claude Code plugins or with npx skills. Includes the free Codebase Readiness assessment, which scores your repo across 8 dimensions of agent-readiness, agent-ready for scaffolding AGENTS.md and regression-aware quality gates, and an engineering bundle for TDD, stacked PRs, and code review
• mastra-meeting-assistant: building a useful meeting assistant agent with Mastra
• claude-code-channels-demo: a custom Claude Code Channel that receives Cal.com webhooks
• buttondown-mcp: an MCP server for the Buttondown newsletter API

• YouTube: weekly videos on AI engineering, agent architecture, and Claude Code
• Blog: deep dives on production AI patterns
• Newsletter: production AI engineering notes for teams shipping AI, every Wednesday

I work with engineering teams shipping AI into production — a two-week Foundation Sprint, then a retainer at the intensity you need. Details at damiangalarza.com/services, or book an intro call.