Corti Launches FactsR: A New Approach to Clinical AI
Corti, a leader in clinical-grade AI infrastructure for healthcare, has introduced FactsR, a real-time agentic reasoning system aimed at clinical consultations. FactsR focuses on ambient documentation, cutting down general AI-driven “note bloat” by 65%, making clinical records more precise and directly relevant to conversations.
This system reduces the need for post-visit edits and converts passive transcripts into actionable clinical intelligence. It sets a clear standard for ambient AI in healthcare and opens the door for real-time decision support during patient care.
How FactsR Stands Apart
Unlike traditional large language model (LLM) pipelines adapted for healthcare after the fact, FactsR runs on Corti’s recursive fact-first reasoning loop. This engine is designed to surface, verify, and organize clinical information in real time as discussions happen.
Delivered via a modular API, FactsR allows developers to embed clinical AI directly into healthcare applications. The result is a safer, leaner, and more reliable AI experience at the point of care.
Most existing ambient solutions generate summaries after consultations by processing raw transcripts with generic models. These often produce lengthy, error-prone notes that clinicians spend hours correcting weekly. FactsR shifts from this passive summarization approach to active reasoning, offering concise and accurate outputs that clinicians can interact with live during consultations.
The Four Key Steps of FactsR
- Listen and Extract in Real Time
FactsR listens during the consultation, continuously extracting structured clinical facts such as symptoms, vital signs, medications, and social history. - Vet and Refine with Specialized AI
Each extracted fact undergoes an AI-driven review and refinement process until it is accurate, consistent, and ready for use—eliminating guesswork and clutter. - Clinician-in-the-Loop
Clinicians can review, accept, or adjust facts as the consultation progresses. Early users report significantly fewer post-visit edits and rarely need to add missing information afterward, maintaining clinician control. - Generate EHR-Ready Notes
Once finalized, FactsR assembles a clean, concise summary free from verbose or irrelevant details. This reduces clinician screen time, keeps decisions within the consultation, and provides audit-ready transparency with timestamps, confidence scores, and links back to the conversation.
Proven Benefits for Healthcare AI
FactsR reduces AI-generated note bloat by 65% while increasing clinical completeness by 13%, according to evaluation on the public Primock57 benchmark dataset. It captures more relevant medical information than traditional ambient scribes, all while keeping clinicians involved in the loop.
A beta tester from a Danish hospital noted, “FactsR goes beyond basic transcription to surface the right clinical facts in real time—exactly what busy clinicians need to focus and cut down on documentation overload.”
Lars Maaløe, CTO and co-founder of Corti.ai, highlights that FactsR embodies a commitment to purpose-built, real-time, and accountable healthcare AI. By breaking conversations into structured facts and validating each one, FactsR transforms ambient documentation into a foundation for real-time clinical intelligence.
The modular API enables healthtech developers to safely and quickly integrate this capability into their applications, enhancing AI’s role from a simple scribe to a trusted clinical collaborator.
For healthcare professionals interested in expanding their knowledge of AI applications in clinical settings, exploring relevant courses at Complete AI Training can be a practical next step.
Your membership also unlocks:
