Mirae, an Oxford University spinout focused on inflammatory bowel disease management, announced on July 28 it has closed a $5.4 million seed round led by Oxford Science Enterprises. The company declined to disclose other participants in the financing.
The startup is attacking a persistent problem in autoimmune care: the information void between clinic appointments. While patients with IBD navigate daily flare-ups, medication side effects and dietary triggers at home, their gastroenterologists often see them only a handful of times each year, making treatment decisions on incomplete snapshots rather than continuous data.
Mirae's answer is a two-sided platform. On one end, patients download a free app—already available on iOS and Android—that uses conversational prompts to log daily symptoms, pulls in data from electronic health records and wearables, and surfaces patterns across lab results, sleep quality, stress levels and physical activity. On the other, clinicians get an AI copilot that digests that patient-reported stream into structured, longitudinal summaries, flagging early warning signs and suggesting treatment adjustments based on clinical guidelines and peer-reviewed research.
"A lot of the variability in care and outcomes comes from the fact that clinicians are working without a full view of what has happened between visits," Anuj Patel, Mirae's co-founder and chief executive, said in an interview. "When you combine experiential data with clinical data, you begin to understand and model disease more effectively and have a path towards true precision medicine."
The approach reflects Patel's own history navigating the intersection of patient experience and clinical workflow. He previously founded Motto Health, a virtual care startup for autoimmune conditions, and held product roles at PatientsLikeMe and iCarbonX. His co-founder, James Nishiyama Finucane, brings operational scale from stints co-founding mPharma and leading remote engineering at Y Combinator-backed GitStart.
Mirae spun out of the University of Oxford's Computational Health Informatics Lab under Prof. David Clifton, who remains an academic co-founder. Companies House filings reviewed by third-party tracker Funding Spotter show Oxford Science Enterprises took a significant stake on April 1. The company formally announced the round in July.
The business model sidesteps the usual digital health playbook of selling software subscriptions or charging patients directly. Instead, Mirae is pursuing shared-savings agreements with health systems and payers, betting that tighter monitoring and earlier intervention will cut hospital admissions and prevent the expensive drug escalations common in IBD treatment. "We do not sell your attention or your data," the company's website states plainly.
Joel Schoppig, a health tech principal at Oxford Science Enterprises who joined Mirae's board in the spring, framed the investment as a bet on specificity. "AI is moving quickly into healthcare, but much of that activity remains broad, generalized, and disconnected from the clinical decisions that determine outcomes," Schoppig said. "Mirae is building for this deeper layer of medicine, where patient context, clinical evidence, and workflow need to come together to support a higher standard of care."
The platform is already being deployed with a large U.S. health system, though Mirae would not identify the partner. That collaboration will feed real-world patient data back into the AI models, helping refine the system's ability to predict flare-ups and recommend adjustments before patients land in the emergency room—a cycle that drives much of the $20 billion annual cost of IBD care in the United States.
Whether the shared-savings model will prove durable remains an open question. Health systems have grown wary of digital health vendors promising cost reductions that never materialize, and demonstrating ROI in chronic disease management can take years. But for a company built on the premise that better information leads to better decisions, Mirae is at least starting with a testable hypothesis.
