Crux Labs has raised $6.0M to build out its platform for evaluating and improving high-stakes AI agents. The round drew backing from Benchstrength, Misneach, Analog Ventures, and Pareto Holdings.
The company's agents automate the full evals lifecycle - capturing expert judgment, creating custom evaluations, monitoring production traces, and generating verified code improvements. It's aimed at teams building customer-facing agents in complex, regulated industries, where a bad output carries real cost.









