Pre Seed · AI & Machine Learning · Cambridge, United Kingdom · Founded 2025 · 1–10 employees
Trismik is a University of Cambridge spinout developing an advanced evaluation and model selection platform for artificial intelligence builders. By applying psychometric testing methodologies such as Item Response Theory and Computerised Adaptive Testing, the company evaluates Large Language Models up to 180 times faster than traditional methods. This adaptive testing approach dynamically adjusts evaluation difficulty, reducing benchmarking costs for developers by up to 95 percent. The company secured a 2.2 million GBP pre-seed funding round in September 2025 to scale its operations and advance its proprietary QuickCompare technology.
No open roles listed right now.