Agents All The Way Down

Are you building generative AI agents but losing sleep over how to ensure they behave correctly? As AI agents proliferate across domains from customer service to code generation, one challenge towers above the rest: how do we systematically test these unpredictable, creative systems for correctness? Traditional testing approaches fall short when your software can generate novel responses you've never seen before. Join Dionny Santiago as he explores the cutting edge of AI agent testing. See real examples across different domains, discover emerging validation techniques, and learn why testing an agent is fundamentally different. But here's where it gets interesting – what if the best way to test an agent is with another agent? This recursive approach is already reshaping quality assurance in surprising ways. Come discover practical strategies for taming your agents and a provocative vision of where AI testing is headed. Spoiler alert: the testers of tomorrow might not be human at all.