​
Introduction

FactsR™ is a real-time agentic reasoning system for clinical consultations. Designed with ambient documentation in mind, FactsR reduces general purpose AI driven “note bloat” by 65 percent, keeping records precise, relevant, and tightly aligned with the actual clinical conversation.

By minimizing post-visit edits and transforming passive transcripts into active clinical intelligence, FactsR sets a new benchmark for ambient AI in healthcare - while unlocking the path to real-time decision support at the point of care.

​
A better brain for clinical AI

Unlike traditional LLM pipelines retrofitted for healthcare, FactsR is powered by Corti’s recursive fact-first reasoning loop - a purpose-built engine designed to surface, validate, and structure clinical knowledge in real time as conversations unfold.

Delivered as a modular API, it enables developers to embed clinical-grade intelligence directly into their healthcare applications - creating safer, leaner, and more trusted AI experiences at the point of care.

Rather than summarizing the full conversation in one go, FactsR breaks the task into three core stages:

  1. Extract – Pull out clinically relevant facts in real time as the conversation unfolds.
  2. Refine – Evaluate each fact using specialized reasoning agents to ensure accuracy, structure, and clinical relevance.
  3. Compose – Build the final note using only validated facts, structured according to standard formats like SOAP.

This architecture doesn’t so much more than improve output quality, it transforms the documentation process.

Corti's system already does an excellent job capturing clinical details accurately, even in natural conversation. But when testing this new innovation, what really stood out was the shift toward structured, recursive fact extraction. It goes beyond basic transcription to surface the right clinical facts in real time - exactly what busy clinicians need to stay focused and cut down on documentation overload.

Feedback from a beta user at Danish hospital

​
Recursive reasoning

Traditional ambient solutions pipe raw transcripts through generic models after the consultation has ended, producing verbose, error‑prone summaries that clinicians spend up to three hours a week correcting.

FactsR reflects a foundational shift in AI system design - from passive summarization to active reasoning. It allows developers to build AI that offers more concise, accurate results that clinicians can interact with live, in consultations and beyond.

The process unfolds in four key stages:

1

Listen and Extract in Real Time

As the consultation unfolds, FactsR continuously identifies and surfaces structured clinical “facts” - such as symptoms, vitals, medications, and social history - while the conversation unfolds, live.

2

Vet and Refine with Specialized AI

Each fact is automatically reviewed and improved through an AI-driven feedback loop. If something is unclear, the system refines it until it is accurate, consistent, and ready to use - no guesswork, no clutter.

3

Clinician-in-the-Loop

Clinicians can quickly review, accept, or adjust facts as they go. Early adopters report far fewer post-visit edits and rarely need to add missing information after the consultation. This design keeps clinicians in control, ensuring that AI augments rather than replaces clinical judgment.

4

Generate EHR-Ready Notes

Once the facts are finalized, the system assembles a clean, concise summary - free from long, verbose summaries or irrelevant content.

​
Measured and published

The innovation behind FactsR has been published together with evaluation results on the public benchmark Primock57 dataset. You can find the research paper here. The evaluation shows that FactsR delivers -

  • 49% reduction in missing clinically relevant content (completeness errors drop from 23.3% to 11.7%)
  • 86% reduction in extraneous detail (conciseness errors fall from 14.3% to 2.0%)
  • Groundedness remains high at 94%, comfortably above clinician reference notes

More of what matters, far less of what doesn't

More Patient Focus
Decisions stay in the consultation, not in hindsight.

Less Screen Time
Spend minutes, note hours, reviewing AI output.

Audit‑Ready Transparency
Every fact carries a timestamp, confidence score, and link back to the conversation.

Healthcare AI that Delivers
Real-time reasoning turbocharges ambient documentation, clinical decision support tools, and more.

Reduce cognitive load without sacrificing control or trust.

To learn more about how to integrate with FactsR™, check out the stream and facts details the in the API reference. You can also read the full announcement blog post here.