Cori Clinical raises $4M to design clinical trials that don't need rewriting

Cori Clinical, a London startup applying AI to the design of clinical trials, has raised $4M (about £3M) in a seed round led by Breega, with Enzo, Heartfelt, Together Ventures and Bynd also taking part. Founded in 2025 by CEO Jorn Jansen Schoonhoven and CTO Patricio Fernández, it is going after one of drug development’s most expensive and least visible problems: trial protocols that turn out to be flawed only after the trial has started.

A clinical trial protocol is the detailed blueprint for how a study will run. Get it wrong — an impractical eligibility criterion, a missed regulatory requirement — and the fix is a formal amendment: slow, costly, and sometimes requiring patients to be re-consented or sites to be re-trained. Those mid-trial amendments are a notorious drain on time and money, and they are exactly what Cori is trying to prevent at the design stage.

How it works

Cori’s platform is an AI workspace that uses retrieval-augmented generation to bring two kinds of knowledge together: a sponsor’s own history from previous trials, and external sources such as trial registries and regulatory guidance from the FDA, the UK’s MHRA and the EMA. From that, it helps teams go from an initial brief to a protocol, a submission package and an operational plan — what the company calls a “living blueprint” that links design decisions to budgets and timelines, so the knock-on effects of a choice are visible before it is locked in.

That retrieval-grounded approach matters in a field where accuracy is non-negotiable. Rather than a general model inventing plausible-sounding trial design, Cori anchors its output in a sponsor’s real prior trials and current regulatory text — the right instinct for a regulated, high-stakes domain. The company has also brought in clinical weight, appointing Dr Emma Ogburn, previously of Lindus Health, as Chief Clinical Officer.

Early traction, and the honest caveat

Cori says it already works with two of the five largest enterprise contract research organisations, and counts the Rotterdam academic medical centre Erasmus MC among its customers, where it supported trials in neurosurgery and neuro-oncology. It estimates its approach could cut trial set-up costs by at least a third, and reports that in an internal evaluation its AI reviewer caught 82% of avoidable protocol errors.

Those last two figures deserve a clear caveat: they are the company’s own modelling and internal benchmarking, not independently verified outcomes. That is not a knock — it is simply where an early-stage company is — but the real test is whether sponsors using Cori genuinely end up filing fewer amendments on live trials. The problem is real, expensive and well chosen, and $4M is enough to go and prove it.

Cori joins a growing cluster of UK companies aiming AI at the machinery of healthcare and life sciences rather than the headlines — from InTouchNow in primary care to the NHS-facing diagnostic work coming out of the UK’s universities. Designing better trials faster is an unglamorous place to apply AI, but if it works, it is one of the most valuable.