PHINEAS
A six-step LLM state machine that forces deterministic CEFR-aligned reading material output from probabilistic models.
Read more
AI Product Manager & AI Operations Lead
Closing the gap between AI demos and AI production, between proofs of concept and operations at scale.
10+ years of complex program delivery at Google and Meta. $25M+ in capital projects at FORTÉ. 100+ station Starline fleet. AI systems shipped to real users at real clients.
A six-step LLM state machine that forces deterministic CEFR-aligned reading material output from probabilistic models.
Read more
A framework for taste-enabled creator agents. Pipeline-plus-judgment architecture, with the agent rejecting its own work when it fails a stylistic bar.
Read more
A multi-agent peer review system built around three-round reconciliation. Reduces grade variance across LLM evaluators from 0.26 to 0.04.
Read more
· 5 min read
An HR consultant asked for pixel art. The discipline that came out of it turned into something I now call agent husbandry.
· 8 min read
Independent LLM evaluators disagree, and averaging their scores hides the disagreement instead of resolving it. Here is a three-round method that surfaces the conflict, reconciles it through structured debate, and converges on an auditable result.