Now · Updated October 2026
What I'm
What I'm
doing now
The short version of right now: governing agents, running a production SaaS, reviewing books, and open to the right role or consulting engagement.
Currently building
- ▸ Turbo Rig — the governed agentic coding rig: builder/reviewer/reserve lanes, cross-family review gate, git-versioned memory, human-only merges.
- ▸ Ancuria — production real-estate intelligence SaaS for Mexico: bilingual AI listing assistant, RADAR market calculators, billing with CFDI invoicing.
- ▸ This site — the canonical identity and evidence hub for my work, including the new expertise section.
Currently researching
- ▸ Cross-family code review — different model families reviewing each other's output; measured at 116 controlled tasks so far (pass rates 71.6% → 89.7%).
- ▸ Agent memory — git-versioned, diffable cross-session context as a first-class engineering artifact.
- ▸ AgentOps — operating metrics for agent fleets: verdict rates, review spend, lane throughput.
Currently reviewing
- ▸ Packt: "Claude Beyond the Prompt" (upcoming, Q4 2026) — the next title in a technical-review run that started September 2020.
Open to
- ▸ Roles: Principal Agentic Engineer · Agentic Platform Engineer · Production AI — remote, US.
- ▸ Consulting: agentic engineering adoption, agent governance design, platform engineering — via Adventure on the Wave.
Latest measured week
141PRs merged
868Gate verdicts
4.21BTokens
One engine, Sept 14–21, 2026 · 5 repositories · source: Adventure Wave Labs