Cohortbridge exists because a Phase III rare disease trial missed its enrollment target — and the patients were right there in the health system the whole time. We’re a seed-stage company based in Durham, NC, building AI-powered eligibility screening for mid-size CROs.
Get eligible patients into trials faster by removing the eligibility screening bottleneck from site coordinators.
Dylan Okoye was a clinical research coordinator at a Duke Health oncology clinical trials unit in Durham when a Phase III rare disease trial missed its enrollment target by 40% in the first six months — not because qualifying patients didn’t exist in the surrounding health system, but because the manual chart review process meant coordinators were screening each patient individually against 38 eligibility criteria using a paper checklist, and 72% of screened patients failed on criterion 34 or 35 after 40 minutes of coordinator time had already been spent on the chart.
The eligibility criteria were computable from EHR data for 80% of cases — the failure was not a patient availability problem, it was a coordination and information retrieval problem. A system that ran the deterministic filters first and surfaced only the uncertain cases for human review would multiply the effective throughput of each site coordinator without adding headcount.
Dylan and co-founder Nneka Obi built a rules engine against Duke’s Epic FHIR API that pre-screened patients against the hard exclusion criteria for two active oncology trials. The tool reduced coordinator pre-screening time from 42 minutes to 9 minutes per candidate and surfaced 23 additional eligible patients in the first three months who had been missed in the manual process.
Cohortbridge now focuses on the two-pass screening model — deterministic rule filters plus NLP for note-dependent criteria — for mid-size CROs serving biotech sponsors, with a CTMS integration layer that delivers outputs into Medidata Rave and Veeva Vault without creating new coordinator workflows.
Get eligible patients into trials faster by removing the eligibility screening bottleneck from site coordinators. Manual pre-screening is slow, inconsistent, and exhausting for coordinators who manage 8–25 screened patients for every enrolled participant. Cohortbridge narrows that ratio by doing the computable work automatically and surfacing only the candidates who genuinely need a trained human’s eye.
We measure success in enrollment weeks recovered and coordinator hours freed — not in features shipped or dashboards built. Every product decision runs through one question: does this help eligible patients get into trials faster?
Cohortbridge is a seed-stage company. We’re not a large enterprise platform and we don’t pretend to be. We’re a focused team of four people — a former CRO coordinator, a health informatics engineer, a CRO operations manager, and a clinical NLP researcher — working with a small number of mid-size CRO partners in the US on oncology and rare disease Phase II-III trials.
That means you get direct access to the people building the system. It also means we’re selective about who we work with: we’re a good fit for CROs running 3–25 active trials with staff of 15–200 FTE who want to reduce screen failure rates without layering new tools on top of coordinators’ existing workflows.
These aren’t statements posted on a wall — they’re the constraints we actually use when making product tradeoffs:
We work with mid-size CROs managing multi-site Phase II and Phase III trials — teams of 15 to 200 FTE with 3 to 25 active trials and annual CRO services revenue between $8M and $120M. We are not built for large global CROs with proprietary patient-matching platforms already deployed, nor for Phase I first-in-human trials with fewer than 20 enrollment targets.
We’re currently onboarding a small number of CRO partners for early access. If you manage Phase II-III oncology or rare disease trials, we’d like to show you what the two-pass model looks like for your specific protocol.