The clinical-grade AI that listens to medical conversations - so clinicians can look at patients, not screens.
Corti is a Copenhagen-based healthcare AI company that builds the models and APIs other medical software runs on. It does not sell a chatbot or a consumer app. It sells clinical-grade infrastructure - speech-to-text, medical coding, note generation, and agentic workflows - tuned specifically for medicine and wired into the systems clinicians already use.
The company was founded in 2016 by Andreas Cleve, Lars Maaloe, Michael Reibel Boesen, and Anders Friis. Its earliest work was unglamorous and high-stakes: sitting alongside emergency dispatchers, listening to 911 and 112 calls in real time, and flagging when a caller might be describing something like a cardiac arrest. That origin - AI as a second set of ears in a moment where seconds matter - still shapes how the company thinks about its products.
Today that capability has moved into the consultation room and the back office. Corti's technology transcribes patient conversations, drafts structured clinical notes such as SOAP notes, assigns medical billing codes, and orchestrates multi-step clinical workflows. The pitch is simple and measurable: clinicians spend hours on documentation for every hour with patients, and Corti's models give a large share of that time back.
Rather than fine-tune a general large language model, Corti built a healthcare-only model called Symphony, trained on more than 1.5 million hours of clinical audio. That bet on specialization is the company's central argument - and, increasingly, its benchmark story.
Clinical-grade AI model trained on 1.5M+ hours of clinical audio, powering applications for EHR vendors, virtual care platforms, and life sciences.
Real-time dictation, conversational transcription, and batch audio processing tuned for medical terminology - reported to beat OpenAI on term accuracy.
Agentic ICD-10 and CPT coding reported to outperform OpenAI and Anthropic by more than 25% on clinical accuracy benchmarks.
Turns consultations into structured clinical notes, reducing documentation time and clinician screen-work.
Tools to build clinical AI agents that coordinate documentation, coding, and decision-support in one workflow.
No-equity, grant-funded program offering founders up to $5,000 in Symphony credits plus clinical and regulatory support.
Corti's whole strategy rests on one claim: for medicine, a model built only for medicine beats a general one. The company reports that its Symphony coding model holds a double-digit accuracy edge over the largest general-purpose labs on clinical coding tasks. Figures below are Corti's reported benchmark comparisons - illustrative, and worth reading as the company's own scoring.
EHR vendors, virtual-care platforms, practice-management systems, insurers, life-science firms, and public-safety dispatch centers across Europe and the US. These customers embed Corti's APIs inside their own products rather than buying a finished app. Through a partnership with BigHand, Corti's copilot is being rolled out to roughly 40,000 frontline workers across 76 NHS trusts.
Clinicians are buried in documentation, coding, and administrative work that pulls attention away from patients and drives burnout. Corti automates the transcription, note-writing, and billing-code assignment around a consultation - reporting reductions of up to 80% in documentation time and around 40% in errors in deployments - so professionals spend more of the visit on care.
The ambient clinical-documentation market is crowded - Nuance/Microsoft's DAX, Abridge, Nabla, Suki, and Ambience Healthcare all chase the same clinician time-savings, while OpenAI and Anthropic push general models into health. Corti's answer is threefold.
Specialization. A model trained only on clinical data, benchmarked directly against the general labs on medical tasks. Infrastructure, not app. Corti sells APIs that other healthcare software builds on, making it a layer rather than a competitor to its customers. Compliance as product. Sovereign hosting, GDPR, HIPAA, and EU AI Act readiness are treated as core features, not afterthoughts.
That last point became a market strategy in 2026. As the EU AI Act's high-risk rules for medical AI approached and MDR certification costs climbed into the hundreds of thousands of euros per device, Corti opened its stack to startups with no-equity credits and regulatory guidance - positioning itself as the on-ramp for the next generation of European health-AI builders even as some larger players retrenched.
Usage- and credit-based access to models and APIs, sold to software vendors, providers, insurers, and dispatch centers who embed the tech.
A machine-learning research culture - CTO Lars Maaloe holds a PhD in ML - that publishes healthcare benchmarks and specialized models.
In-house clinical and regulatory expertise to navigate EU AI Act, MDR, HIPAA, and data-residency demands - a moat in a slow, high-trust market.
Andreas Cleve, Lars Maaloe and co-founders start Corti to help machines understand medical conversations.
Corti builds real-time decision-support for emergency call centers, detecting critical conditions during calls.
Prosus Ventures and Atomico co-lead a round to scale healthcare AI and expand into public safety.
Corti introduces its flagship clinical-grade model and broadens its developer API platform.
New speech-to-text and coding models post accuracy edges over OpenAI and Anthropic; a no-equity accelerator opens.
Co-led by Prosus Ventures and Atomico, with Eurazeo, EIFO, and Chr. Augustinus Fabrikker participating - to scale healthcare AI and expand into public safety.
Earlier rounds from backers including Nordic Makers, Sunstone, and Atomico, bringing total funding above $94M across the company's history.
"Corti's technology can make an important impact on our overall wellness - the aim is not to replace clinicians, but to give them superpowers."— Andreas Cleve, Co-founder & CEO (paraphrased from interviews)
Corti builds clinical-grade AI models and APIs - speech-to-text, medical coding, clinical note generation, and agents - that healthcare software and providers use to automate medical documentation and support clinicians.
Symphony is Corti's flagship healthcare AI model, trained on over 1.5 million hours of clinical audio, with specialized versions for speech-to-text and medical coding that it reports outperform OpenAI and Anthropic on medical tasks.
EHR vendors, virtual-care platforms, insurers, life-science firms, and public-safety dispatch centers across Europe and the US - reaching roughly 100 million patient interactions a year, including NHS trusts via a BigHand partnership.
Corti has raised more than $94M in total, including a $60M Series B in September 2023 co-led by Prosus Ventures and Atomico.
Corti is headquartered in Copenhagen, Denmark, with offices in London and New York.