About

Tony Williams

Tony founded TrendLab Analytics after 17+ years in healthcare finance, underwriting and analytics, building data solutions for Blue Cross Blue Shield, Aon Hewitt, Willis Towers Watson and OneDigital. He has a decade of production Python and SQL (Snowflake, BigQuery, Teradata, PostgreSQL), and advanced Tableau.

Today he leads enterprise AI adoption at a Fortune 500 financial institution and works daily in the agentic toolchain: Claude Code, Codex CLI, Pi, GitHub Copilot CLI, Gemini CLI and local LLMs.

How the analytics background carries into AI work

Analytics work is reconciling numbers, testing edge cases and saying plainly what the data does and does not show. Evaluating an AI agent is the same discipline applied to a system that answers differently every time. Tony built TrendLab's agent evaluation engine (sealed run records, two-provider judging, critical tests reported apart from averages) and runs it every day on TrendLab's own agents. He also built an experimental agent harness as a research lab, to learn from the inside where planners, verifiers and model routing help and where they only add cost. TrendLab does not sell a harness; it configures proven ones and measures them.

The analytics and automation results on this site come from analytics work; they are not AI evaluation results. TrendLab's evaluation work is presented with its own evidence and availability on the AI evaluation pages.

How TrendLab is organized

TrendLab Analytics & SuperIntelligence is one owner-operated firm. Its three areas (analytics and automation, AI engineering, AI evaluation) are kinds of work, not separate companies. We only call an evaluation independent when we had no hand in building or advising on the agent, and our price is fixed before testing starts, so the results don't change what we're paid. Brand names have nothing to do with it. Details: independence rules.

TrendLab Analytics is based in Atlanta, Georgia, serving clients nationwide.