Voice AI Meetup: Promptable Speech Models

​We built Universal-3 Pro so you can tell the model what matters, before it even listens. Whether you're building voice agents, medical transcription tools, or real-time call intelligence apps, prompting can be your unlock. And we're bringing the people who built it to show you exactly how to use it.

​Join AssemblyAI's technical team for a night where we'll demo our async and soon-to-be launched Universal-3 Pro Streaming models, dig into the real stuff (messy audio, turn detection, entity accuracy, speaker diarization), and open the floor to your burning questions.

​Prompt It. Shape It. Ship It. — A Universal-3 Pro & Universal-3 Pro Streaming Deep Dives

​🔥 Burning Topics we'll cover (builders ask...we listen):

• ​Entity accuracy — credit cards, phone numbers, emails, names. U3 Pro Streaming nails these at sub-300ms latency.

• ​Speaker diarization via prompting — skip the speaker_labels param, let the model figure it out from context.

• ​Turn detection for voice agents — Built-in turn detection optimized for voice agent use case.

• ​Keyterms vs. prompt — you can't use both at once. We'll tell you which one win for your use case.

• ​PII redaction and verbatim control — one prompt, zero post-processing pipelines.

​Agenda

• ​5:30–6:30 pm — Networking + Happy Hour

• ​6:30–7:00 pm — Live Demo — U3 Pro Async & Streaming

• ​7:00–7:45 pm — Speaker Panel + Q&A

• ​7:45-8:30 pm — Networking

​🔗 assemblyai.com | docs | Try U3 Pro →

​​​​Limited Capacity

​​​​Seats are limited. Request an invite to join the community building the future of voice AI!

​​By signing up for this event, you opt in to receiving AssemblyAI marketing emails. You may unsubscribe at any time. By attending this event, you acknowledge that photography and video recording may take place; if you do not consent, please contact the event host.