Enterprises are increasingly using programmatic verifiers, or software programs, to grade the performance of AI agents, rather than relying on another AI as a judge. This approach is gaining traction as businesses deploy agents for more critical operations. Scale AI's director of general agents, Brad Kenstler, noted that enterprises are becoming more sophisticated in evaluating agents, and Scale is discussing these verifiers with clients such as Mayo Clinic and Cengage.
Related Articles
Don't miss out on breaking stories and in-depth articles.