About Spec27
Spec27 is a Safe Intelligence product designed to help teams validate and monitor language model-based AI Applications and Agents. The foundational technology is based on sophisticated validation and verification techniques that generate test cases in efficient rigorous ways.
Spec27 grew out of many customer conversations, all focused on one question: ‘How can we get on top of validating language-model-based AI agents?’ Validating AI is extremely challenging, and the stakes rise rapidly when it is used in applications with little or no human in the loop. That set us on a mission to deliver best-in-class stress testing for AI applications. Spec27 is the result.
Safe Intelligence was founded with the vision of creating radically improved capabilities for validating AI systems and technology. The company’s aim is to advance safety, security, reliability and trust in environments where computers control or influence critical applications. Find out more at SafeIntelligence.AI.
Founder
Steven Willmott is Founder and CEO of Safe Intelligence. He holds a doctorate in AI from EPFL, has published extensively in AI and distributed systems, and previously co-founded 3scale, later acquired by Red Hat, where he went on to lead API infrastructure initiatives.