> ## Documentation Index
> Fetch the complete documentation index at: https://docs.corti.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# FactsR™ Extraction

> Learn more about ways to use FactsR to extract facts

<Info>Learn about the differentiating factors and research fuelling [FactsR here](/textgen/factsr). This guide is a more hands-on API guide.</Info>

## Purpose and Behaviour of FactsR

Rather than summarizing the full conversation in one go, FactsR breaks the workflow into four core steps:

<Steps titleSize="h4">
  <Step title="Extract">The Corti LLM pulls out clinically relevant facts.</Step>
  <Step title="Refine">Each fact is automatically evaluated along medically-relevant criteria and improved through an agentic feedback loop to ensure accuracy, structure, and clinical relevance.</Step>
  <Step title="Human Review">Clinicians can quickly review, discard, or adjust facts. This design keeps clinicians in control, ensuring that AI augments rather than replaces clinical judgment.</Step>
  <Step title="Compose">Generate the final, EHR-ready document using only validated facts, structured according to standard formats like SOAP or other formats facilitated via custom templates.</Step>
</Steps>

## How to Extract Facts with FactsR

Facts can be extracted in both a real-time and asynchronous workflow.
<Tip> Click on the cards to learn more...</Tip>

<Columns cols={2}>
  <Card title="Real-time" icon="radio-tower" href="/textgen/facts_realtime">
    Extract facts in real-time as the ambient conversation unfolds.
  </Card>

  <Card title="Asynchronous" icon="list-tree" href="/textgen/facts_async">
    Extract facts via REST, independent of a WebSocket audio stream, from a text input.
  </Card>
</Columns>