Enhance Your Transcripts with Our Speaker Identification API

We can identify over 25k speakers with a 99% accuracy for identifying speakers across a diverse set of voice domains.

CorporateUS GovernmentJournalistsUK House of ParliamentCelebrities

Try it out yourself

Generic diarization

[0:00:01] SPEAKER_01: Good morning, everyone. Thank you for joining our Q2 earnings call.

[0:00:05] SPEAKER_02: Good morning. We’re excited to share our financial results with you today.

SpeakerMap diarization

[0:00:01] CEO, Maria Lopez: Good morning, everyone. Thank you for joining our Q2 earnings call.

[0:00:05] CFO, David Kim: Good morning. We’re excited to share our financial results with you today.

Our Offerings

Unlock clarity, precision, and attribution in audio processing with our suite of services. From speaker identification to transcription and diarization, we ensure every voice is heard and understood.

Speaker Identification

Provide an audio file or an utterance from a single speaker, and we'll deliver a diarized transcript of the audio along with the speaker's name, company, and title.

Disambiguation & Search

Submit a name, voice sample, or other known details and we'll return the canonical name along with all known aliases, helping you clarify and connect speaker identities.

Transcription & Diarization

We provide highly accurate transcriptions and speaker diarization to ensure clarity and ease of understanding in your audio documents.

How Speaker Identification Works

Submit any audio—from earnings calls to congressional hearings—and get a comprehensive speakermap complete with speaker IDs, aliases, and enhanced profiles through our API.

Full Audio

Input

Provide a full-length recording—be it a congressional hearing, earnings call, or justice department hearing. If it matches our domain set, our API will identify the known speakers.

Or, if you provide a diarized transcript, our API will return a speakermap which maps the labels in your diarization output to the recognized speaker.

SpeakerMap
API Service

Speaker Map

Output

[
  {
    "SPEAKER_1": {
      "predicted": "john-smith",
      "aliases": ["jon-smith", "j-smith"],
      "confidence": 0.2788,
      "profile": {
        "linkedin_url": "linkedin.com/in/john",
        ...fullProfie
      }
    }
  },
  {
    "SPEAKER_2": {
      "predicted": "jane-smith",
      "aliases": ["j-smith"],
      "confidence": 0.2788,
      "profile": {
        "linkedin_url": "linkedin.com/in/jane",
        ...fullProfie
      }
    }
  }
]

Audio sample

Input

Submit a brief 20-second clip of a single speaker and our API will deliver a complete profile, including aliases and a confidence score.

SpeakerMap
API Service

Speaker Profile

Output

{
  "predicted": "jane-smith",
  "aliases": ["j-smith"],
  "confidence": 0.2788,
  "profile": {
    "linkedin_url": "linkedin.com/in/jane",
    ...fullProfie
  }
}

Voice Domains

Our current databases encompass federal government, corporate and entertainment domains. We are expanding our database to include podcast, and state and local governments and international speakers.

You can run a test search within each domain, and we will provide both the canonical names and aliases of the individual you searched for.

Corporate

20k

Unique Speakers

6k

Speaker Aliases

9k

Companies

27k

Earnings Calls

Congress

907

Congress Members

321

Member Aliases

7k

Legislative sessions, committee hearings

Celebrities & Journalists

Q4 2024

Frequently Asked Questions

Find answers to frequently asked questions about our speaker identification services.