Assessment Labs

About Assessment Labs

For twelve years, I assessed executives for multinationals and professionals across United Nations agencies, where every finding had to be grounded in the candidate’s own words. In January 2023, I began using LLMs to do that reading. In blind comparisons across 100+ UN candidates, my AI-assisted assessments matched the independent consensus of other assessors about 95% of the time. Assessment Labs turns that method into an engine. It analyzes language against the client’s own criteria and drafts the assessment report. Every finding is tied to a verbatim quote that the engine verifies against the transcript, making the output reliable and auditable. We start with assessment firms and internal talent teams that already conduct the interviews but cannot read them consistently at scale. The same engine can then identify behavioral outliers for lenders and platforms. Why now: the market is automating the wrong half. AI interviewers ask the questions, while candidates increasingly have reasons to walk away; meanwhile, the EU classifies AI systems used in employment as high-risk. We automate the reading—not the human interaction—and keep the human in the decision.