Spec27 at Google DeepMind Startup Session

On 11 June, we joined the Google DeepMind Startup Sessions at the Ministry of Sound in London – a morning of talks, demos, and live pitches bringing together AI startups and the Google DeepMind community.

The lineup included a talk by Dr Raia Hadsell, a recap of Google’s annual developer conference, and a pitch showcase featuring around ten selected startups. We were glad to be one of them, presenting Spec27, our product for validating how AI agents behave before deployment and at runtime.

Raia Hadsell on the latest Google AI advances

Dr Raia Hadsell, presenting Google AI latest advances at the Startup Sessions.

The standout moment was Raia Hadsell’s talk, which covered several of the most active areas in AI research today: embodied intelligence, world models, high-speed diffusion models, and Gemma, Google’s family of open models for developers building AI applications.

As models become more capable and complex, the hard question shifts from “Can it do this?” to “Can we trust it to?”

That is precisely the gap we are working on at Safe Intelligence. Spec27 helps teams define, test, and validate the behaviour of AI agents and AI-powered applications, so the move from an impressive demo to a production system stops being a leap of faith.

This seemed to resonate with many attendees who came to speak with us after our CTO, Alessio Lomuscio, took the stage to present Spec27. It was encouraging to see so many teams not only building with AI, but also thinking seriously about validation from the outset.

Alessio Lomuscio, CTO of Safe Intelligence, presenting Spec27 at the Startup Sessions.

Thank you

A big thank you to Umar Noor, Oksana Stowe, and everyone else at the Google DeepMind AI Studio team for putting the day together. We were glad to be part of it and look forward to staying connected with the teams and community we met there.