Nate Brown

I lead quality engineering teams, and the part of the job I care about most is developing the people on them. Over 17 years I’ve worked on telecom test platforms, a retail returns software platform, and biometric identity systems used by U.S. law-enforcement agencies.

At IDEMIA I inherited a 16-person QA team with chronic turnover. I grew it into an organization of 25+ across QA and customer success, led through team leads and managers. We cut escaped defects about 70% in one year and another 30% the next. We moved quality earlier with acceptance-criteria reviews before development and automated quality gates in the build, and replaced hand-built reporting with dashboards leadership could read for themselves. I also instituted a 40-hour week for my team. Once escaped defects fell and throughput rose, engineering leadership extended it to the development teams.

My AI work is about evaluation: deciding whether AI output can be trusted before it ships. At IDEMIA I built automated evaluation that used a large language model (LLM) as a judge, scored against expert-reviewed golden datasets. Releases were blocked if the model leaked personal data or refused requests it should have answered. I’ve also run inter-rater agreement (kappa) studies on how consistently people classify defect severity. That’s the same question you have to ask of any judge, human or model.

Through my consulting practice, Praxis Assurance, I help companies plan testing for legacy-system migrations. I’m looking for a remote Director-level role in quality engineering or AI evaluation, leading people and owning whether what ships can be trusted.