University of Cambridge spinout Trismik raises funding to target market worth billions

The University of Cambridge spin-out is developing an adversarial platform designed to revolutionise the large language model testing field.

The testing platform is said to ensure model accuracy, safety and alignment with societal values, enabling AI professionals to build applications with fast deployment and mitigating risk.

Co-founder and CEO Rebekka Mikkola is backed by Professor Nigel Collier, the company’s Chief Scientist and Professor of Natural Language Processing at the University of Cambridge, and by CTO and co-founder Marco Basaldella.

The company, now scaling with enterprise partners, is based at the Future Business Centre, Kings Hedges Road, Cambridge and was only incorporated in January.

Trismik describes itself as “an early-stage spinout founded by technology geeks with roots in Cambridge University, Salesforce and Amazon. With over 38 years of research experience, 200+ published papers, and more than 12,000 citations, we’re transforming traditional Large Language Model evaluation. Our breakthrough technology assesses models up to 180x faster, dramatically increasing efficiency and accelerating project turnaround.”

Trismik’s adaptive evaluation platform is said to tackle real-world risks like hallucinations and bias, delivering results up to 180x faster than traditional methods.

With early support from the Technology Investment Fund (TIF) and as a recent graduate of the Founders at the University of Cambridge START 2.0 accelerator programme, this technological solution has grown from a research insight into a venture shaping the future of responsible AI.

Trismik’s solution is said to be up to 80 times faster than anything else on the market while offering up to 95 per cent cost savings. It is said to already be working with a major telco.