Delfa

Delfa builds AI agents and an accompanying software platform that clinical research sites use to recruit and retain trial participants. The system contacts prospective participants by voice call, chat, SMS and email, captures leads, runs prescreening conversations, schedules and confirms visits, and sends reminders across study visits, with multilingual support. Surrounding the agents, Delfa offers a unified inbox for participant conversations across channels, a prioritized task list ordered by site priorities such as critical studies and upcoming deadlines, real-time recruitment funnel reporting from lead to enrolled participant, patient profiles, and database querying to match existing participants to open studies [0][1][2].

The company positions the product as an AI-native Patient Relationship Management (PRM) system that sits alongside, rather than inside, a site's clinical trial management system, on the premise that CTMS software was not designed for recruitment workflows. Delfa syncs recruitment data back into the site's CTMS and CRM to remove manual data entry [2]. Agents are built custom per study by Delfa's team, which the company says includes medical doctors and AI engineers, and are governed by "agent operating procedures" (AOPs) that set guardrails per workflow, with human clinical review and quality/safety checks applied to agent executions [0][2]. Onboarding is run by the founders in three stages: a one-hour kickoff to map workflows and integrate the CTMS, roughly two weeks to design and deploy custom agents, and ongoing monitoring via shared Teams or Slack channels [0].

Coverage of the seed round describes voice AI as the company's wedge into a larger ambition: turning collected recruitment data into operational insight (for example which sites suit which studies) and building an AI-native CRM for patient recruitment [4]. A 2026 job posting indicates the company is also developing products aimed at biopharmaceutical companies in addition to sites [7].

Founding story

CEO Tom Farrand, a biochemical engineering graduate of the University of Bath (Master's, 2017), worked for three years at IBM in London on computing solutions for UK hospitals, then spent four years in machine-learning-focused startup roles as an engineering leader and head of product. He met CTO Gustav Bredell while both were at Entrepreneurs First as prospective founders in September (per the article, the September preceding the August 2025 story). Bredell holds a PhD from ETH Zurich in deep learning and medical image analysis and interned at Roche working on foundation models for pathology images. The idea originated in Bredell's observation that technically trained colleagues at hospitals and research facilities spent their days in Excel cold-calling people to recruit them into trials. The pair spoke with roughly 300 contract research organizations and found sites overwhelmed by manual admin and dozens of disconnected tools, concluding that recent improvements in voice AI made automating this feasible [4].

Business model

Delfa sells software and custom-configured AI agents to clinical research sites and site networks, deployed against their studies and integrated with their CTMS/CRM systems. Delivery includes a services-like onboarding motion (workflow mapping, agent design and deployment over roughly two weeks, ongoing monitoring and strategic check-ins), and the company reports recurring revenue [0][2][7].

Sources report seven-figure annual recurring revenue growing 25% month over month, indicating a recurring subscription-style model, but no pricing structure is disclosed [7].

Traction

At the time of the August 2025 seed round Delfa was powering more than 50 live trials, worked mostly with U.S. customers, counted site networks including Flourish, Elevate and Re:Cognition as customers, and was in talks for pilots with two top-20 global pharmaceutical companies [4][5]. A later job posting states deployment on 300+ trials across multiple conditions, seven-figure ARR growing 25% month over month, and an expectation of supporting recruitment of thousands of participants in 2026 [7]. Published case studies cite 101 patients randomized in about five weeks with $42,000 of labor effort automated at a U.S. weight-management site network, 7% of leads randomized and 66% of self-scheduled patients reached in a cardiovascular study, and 50% of a site target enrolled in a week with 20 hours per recruiter saved weekly in a migraine study [0][2].

Latest developments

A November 2025 partnership with Pantheon Clinical Research extended Delfa's PRM and AI agents across that site's recruitment operations [3]. The company's news page also lists an article dated 05.08.2026 titled "How to Validate AI in Clinical Research" [3]. Recent hiring materials cite 300+ trials, seven-figure ARR growing 25% month over month, plans to support recruitment of thousands of participants in 2026, and new products for biopharmaceutical companies [7].