Cambridge-based startup Trismik has secured £2.2 million in pre-seed funding to advance its efforts in making artificial intelligence testing more efficient and reliable. The company, a spinout from the University of Cambridge, is developing tools aimed at improving the safety and reliability of large language models (LLMs), which are increasingly integrated into products and decision-making processes.
Testing LLMs for safety, reliability, and performance presents ongoing challenges due to current methods being slow, manual, and inconsistent. Trismik's approach uses adaptive testing and automatic scoring to assess models on factors such as factual accuracy, bias, and toxicity. This system dynamically selects test cases that are most informative, reducing the number of required data points while maintaining high reliability and supporting faster development cycles.
“AI is no longer just generating answers, it's shaping decisions, products and lives. If we want trustworthy AI, we need to treat evaluation as seriously as we take training. Trismik aims to lead that charge by giving AI engineers the tools to test with precision, act with confidence and build with integrity,” said Nigel Collier, Professor of Natural Language Processing at the University of Cambridge and co-founder and Chief Scientist at Trismik.
Collier began his career in machine translation using neural networks during the 1990s but has since shifted focus toward ensuring AI systems function as trusted partners rather than risks. His interest in whether AI could be evaluated similarly to humans inspired Trismik’s method for adaptive evaluation.
Trismik was co-founded by Collier alongside Rebekka Mikkola—an experienced founder with a background in enterprise sales—and later joined by Marco Basaldella as CTO in 2025. The founding team combines scientific research experience with commercial expertise.
Dr Christine Martin, Head of Ventures at Cambridge Enterprise, said: "Trismik exemplifies Cambridge’s continued contribution to global AI development with the team combining world-class academic credentials and practical industry experience that has given them the unique authority to define how AI capabilities should be measured. By solving a pivotal challenge in AI adoption, Trismik is positioned to drive trust at scale - we’re excited to support their journey to market."
The recent funding round was led by Twinpath Ventures with participation from Cambridge Enterprise Ventures, Parkwalk Advisors, Fund F, Vento Ventures, and angel investors from Ventures Together.
More details can be found on the Cambridge Enterprise website.