Almanac Health Raises $10M Seed to Scale Research-Validated Clinical AI — Founded by the Pioneer of Clinical Retrieval-Augmented Generation
AI / Machine Learning | 3 min read
Almanac Health (Boston; a research-validated clinical AI platform and evidence-based clinical decision support company; Founder and CEO Cyril Zakka, MD), has raised $10 million in a seed round led by F-Prime, with participation from General Catalyst and Lightspeed Venture Partners. General Catalyst previously led the company's pre-seed round, with participation from Soma Capital. Total funding now stands at nearly $12 million. The financing will accelerate Almanac Health's mission to bring safe, evidence-grounded AI to clinicians and health systems — building a unified clinical AI platform that delivers specialist-grade knowledge to every clinician at the point of care, integrated with existing EHR systems, governed by institutional controls, and grounded in peer-reviewed evidence. The platform is currently undergoing rigorous clinical validation in academic medical centre settings.
"There is no shortage of technologies promising to transform medicine. But few are grounded in real evidence, built with integrity and deserving of trust from the people and organisations using them. Almanac Health's goal is to build clinical AI held to that standard. Validated through research, with incentives aligned with clinicians and patients. We're focused on getting that right."
— Cyril Zakka, MD, Founder and CEO, Almanac Health
Cyril Zakka, MD — The Clinician Who Invented Clinical RAG and Led Health AI at Hugging Face
Almanac Health is built on an unusually deep scientific foundation. Cyril Zakka, MD worked at the intersection of medicine and AI across clinical practice, academic research, and industry — earning his MD before conducting machine learning research at Stanford, where he developed the first retrieval-augmented generation (RAG) system for clinical medicine. That work — published in NEJM AI and recognised as one of the journal's most-cited papers — established the foundational approach now widely adopted across clinical AI for grounding generative AI in trusted medical knowledge rather than broad unverified datasets. Zakka subsequently led health AI at Hugging Face, the largest open-source AI platform, before founding Almanac Health. The platform operates free from pharmaceutical advertising, ensuring that clinical recommendations are governed entirely by evidence and institutional priorities rather than commercial incentives.
"Cyril is one of the rare founders who has both the clinical training and the technical depth to build AI that clinicians will actually trust as well as the discipline to validate it before deploying it. Almanac Health represents what we believe the next generation of healthcare AI should look like: grounded in research, governed by institutions, and built with incentives that put clinicians and patients first."
— Carl Byers, Partner, F-Prime
Key Takeaways
- • Almanac Health (Boston; Founder and CEO Cyril Zakka, MD; research-validated clinical AI platform; evidence-based clinical decision support) has raised $10M seed round — announced 23 April 2026. Led by F-Prime (Carl Byers, Partner); General Catalyst and Lightspeed Venture Partners participating. General Catalyst previously led pre-seed (Soma Capital participating). Total funding: ~$12M. Funds: grow team across clinical medicine/AI research/health systems infrastructure; advance R&D with focus on privacy, reliability, and rigorous evaluation before deployment.
- • Almanac Health platform: unified clinical AI for clinical decision support — brings specialist-grade knowledge to every clinician at the point of care across medical specialties. Integrates with existing EHR systems. Governed by institutional controls. Grounded in peer-reviewed evidence. Free from pharmaceutical advertising. Currently undergoing rigorous clinical validation in academic medical centre settings. The platform is designed to translate frontier AI intelligence into the critical, everyday decisions made by clinicians — without disrupting established workflows.
- • Cyril Zakka, MD — founder background: MD; Stanford machine learning research; developed the first retrieval-augmented generation system for clinical medicine — published in NEJM AI as one of the journal's most-cited papers; established the foundational approach now used by most clinical evidence AI products. Subsequently led health AI at Hugging Face (largest open-source AI platform) before founding Almanac Health. His RAG approach anchors generative AI in trusted medical knowledge rather than broad, unverified datasets — providing the clinical accuracy and reliability that clinicians require to trust AI-generated recommendations at the point of care.
- • What distinguishes Almanac from the crowded clinical AI market: (1) Peer-reviewed research foundation — the RAG methodology is published and validated, not proprietary black box. (2) Pharmaceutical advertising-free — recommendations governed entirely by evidence and institutional priorities. (3) Institutional governance — hospitals maintain authority over how AI-generated intelligence is used within their environments. (4) Validation-first — the platform is undergoing rigorous clinical validation in academic medical centres before broad commercial deployment. (5) Founder credibility — rare combination of MD clinical training, Stanford ML research, and Hugging Face industry experience in a single founder.
- • Investor framing — F-Prime's Carl Byers describes Almanac as "what the next generation of healthcare AI should look like: grounded in research, governed by institutions, and built with incentives that put clinicians and patients first." The broader investor thesis: as healthcare systems grapple with clinician burnout and the need for more efficient data-driven care delivery, clinical AI tools that lack evidence, trust, or institutional governance are facing growing resistance from healthcare leaders. Almanac's validation-first, research-grounded, pharma-advertising-free model is positioned as the credible alternative — building the standard for clinical AI with the same rigour that the medical profession applies to new treatments before they reach the bedside.