Speaker Identification with Pre-Enrolled Voices using Speechmatics

Speaker diarization tells you who spoke when. Speaker identification goes one step further — it tells you who that speaker actually is. In this demo, we show how Speaker ID with pre-enrolled voices works using Speechmatics, enabling you to reliably identify known speakers in real time or batch transcription workflows. You’ll see how this is useful for: Voice agents with returning users Contact centres and authenticated callers Healthcare and clinical conversations Any multi-speaker scenario where identity matters, not just turns This capability builds on Speechmatics’ industry-leading diarization, adding the option to match speakers against known voice profiles — even in noisy, real-world audio. Chapters: 00:00:00 - Introduction to Speaker Identification and Session Overview 0:00:57 - Speaker Enrollment Process and Python Implementation 0:02:11 - Executing the Enrollment Script and Using the Speaker Identifier 0:02:50 - Configuring Real-Time Speaker Identification 0:04:41 - Demonstration and Results of Real-Time Speaker Identification 🔗 Useful links: Speaker ID example code (GitHub): https://github.com/speechmatics/speechmatics-academy/tree/main/basics/09-voice-agent-speaker-id Speech-to-Text overview: https://www.speechmatics.com/speech-to-text Voice AI & Voice Agents use cases: https://www.speechmatics.com/use-cases/ai-voice-agents Try it yourself in the Portal: https://portal.speechmatics.com 💡 Why Speaker ID matters Pre-enrolled speaker identification unlocks more intelligent Voice AI — enabling personalization, authentication, analytics, and better downstream automation when paired with LLMs and agent frameworks like Vapi. If you’re building production voice systems and need accuracy, low latency, and reliability at scale, this is how Speechmatics helps you go beyond basic transcription.