Streams endpoint

Real-time conversational clinical intelligence

Interested in building a real-time intelligence or clinical decision support solution? Look no further than the Corti AI Streams API: Real-time, bidirectional WebSocket Secure (WSS) communication with the Corti AI platform for live transcript generation and clinical fact extraction (FactsR™).

Connecting from a browser? Never expose client_id/client_secret or a full-scope access_token to the frontend. Issue a limited-scope token with scope="openid streams" from your backend so the token can only be used against this WebSocket endpoint. See Limited-scope credentials for streaming APIs.

Delivered as a modular API, /streams provides a stateful interaction framework for developers to embed clinical-grade intelligence directly into their healthcare applications, creating safer, leaner, and more trusted AI experiences at the point of care.

Depending on your use case, the streams endpoint can be used for real-time conversational transcript, fact extraction, or both! See the full API specification here.

---

Using the API

1. Initiate a /streams by creating an Interaction, which will return a web socket URL along with the interactionId.

2. Connect to the web socket and set your configuration:

   Parameter	Description
   primaryLanguage	Spoken language to be transcribed
   isDiarization	Enable speaker separation (most useful on single channel audio)
   isMultichannel	Enable multichannel audio.
   participants	Assign speaker roles for audio channels. Must be used in conjunction with multichannel: true.
   mode.type	Define facts or transcription depending on the desired real-time output
   mode.outputLocale	Output language for extracted facts (required with "type":"facts")

   See detailed configuration options here
3. Once config is accepted, begin sending audio packets.

4. Receive transcripts every ~3 seconds and facts every ~60 seconds (standard response times can adapted for custom response times).

5. Send the end message to close the audio stream.

---

Features

Click on the cards to learn more…

Languages

Corti speech to text is specifically designed for use in the healthcare domain. A tier system has been introduced to categorize functionality and performance that is available per language and endpoint. Languages in the Enhanced and Premier tiers have the utmost functionality and recognition accuracy - they’re the ones recommended for dictation use.

Audio Configuration

With support for mono or multi-channel audio, with live transcoding and a variety of file formats to choose from, don’t let the complexities fo audio capture and processing inhibit opportunities for real-time intelligence. Read more about our recommendations and best practices.

FactsR

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.

Formatting

Speech to text can be used to create a verbatim transcript of the audio; however, some content is not documented in the same manner as it is verbalized. The formatting features assures that key information (dates, numbers, measurements, etc.) are output as expected in the transcript.

Diarization

Diarization is the process of segmenting an audio recording by speaker, assigning portions of speech to distinct identities (e.g., “Doctor,” “Patient”). This enables accurate transcription, attribution, and analysis of multi-speaker clinical conversations, but is not required for effective AI scribing or workflow speech-enablement.

Audio Events

Real-time events during audio streaming about quality and speech activity, intended to notify integrator of audio health degradation, periods of silence, or other events that could support application behavior or user warnings.

Please contact us for more information or help.