Skip to content
YC Root AccessYC Root Access

Tavus: The AI Human Platform

Tavus is building real-time AI humans — systems that can see you, hear you, and respond with natural expression, emotion, and context. What began as personalized video has grown into a full platform used by companies from startups to the Fortune 10. The team recently raised a $40M Series B to advance this vision, introducing PALs: agentic AI humans that can perceive, reason, and act on their own. In this conversation with YC’s Diana Hu, founders Hassaan Raza and Quinn Favret share how they made the leap from generative video to real-time AI humans, the foundational models behind rendering and perception, and why they believe AI humans will become the next major interface for work and communication. Learn more about Tavus at https://www.tavus.io. Chapters: 00:24 – From Personalized Video to AI Humans 01:18 – Why Real-Time Matters 02:36 – How AI Humans See, Hear, and Respond 04:05 – Introducing PALs: Agentic AI Humans 05:42 – The Foundational Models Behind Tavus 07:28 – Building Emotion, Expression, and Context 09:10 – Use Cases From Startups to the Fortune 10 11:00 – Raising the $40M Series B 12:52 – The Future: AI Humans as the Next Interface

Diana HuhostHassaan RazaguestQuinn Favretguest
Nov 13, 202515mWatch on YouTube ↗

At a glance

WHAT IT’S REALLY ABOUT

Tavus builds real-time AI humans as next computing interface platform

  1. Tavus builds real-time “AI humans” that can see, hear, and respond with humanlike expression across video, audio, and text.
  2. The company evolved from early lip-sync personalized video (pre-ChatGPT) into an SDK/API platform after choosing to be an AI research lab rather than an AI sales tool.
  3. Their core technical thesis is that believable AI humans require both high-fidelity rendering and contextual perception of facial expressions, gestures, and conversational timing.
  4. Tavus customers range from startups to Fortune 10 companies using the tech for AI employees in training, healthcare, and go-to-market functions.
  5. The founders address job-replacement concerns by emphasizing improved experiences where automation already exists and expanding access in areas like therapy where the alternative is often nothing.

IDEAS WORTH REMEMBERING

5 ideas

Real-time latency is a prerequisite for “feeling human.”

They argue natural conversational back-and-forth depends on very fast turn-taking (on the order of hundreds of milliseconds), otherwise the interaction feels machine-like regardless of model quality.

Humanlike agents require perception, not just a photorealistic face.

Tavus frames the “yin and yang” as rendering plus contextual perception—reading expressions, gestures, and how something is said—to model the full meaning of human communication.

Tavus’s pivot was an identity decision: research lab vs. vertical app.

After early traction selling personalized video to sales teams, they deliberately churned customers to focus on foundational models and an SDK/API, aligning with the team’s technical DNA and long-term vision.

PALs aim to move AI from “command line” to a more universal interface.

They compare today’s AI UX to early computing and position AI humans as the GUI-like shift that lets non-technical users interact naturally via calling, texting, and proactive assistance.

Agentic AI humans are positioned as coworkers, not chatbots.

The demo emphasizes proactive task execution (e.g., drafting an email after a schedule change) and continuous availability, aiming to make the experience feel like collaborating with a colleague.

WORDS WORTH SAVING

5 quotes

We build AI humans. We're an AI research lab that focuses on teaching machines the art of how to be human.

— Hassaan Raza

At like, you know, an, a great response back and forth happens in, like, less than 200 milliseconds.

— Hassaan Raza

Human conversation is an art, it's a dance… we're doing a waltz… and machines are in the corner doing the robot.

— Hassaan Raza

The alternative isn't a human, it's nothing.

— Hassaan Raza

The only thing that matters is momentum.

— Quinn Favret

Real-time multimodal AI humansLow-latency conversational interactionPerception + rendering foundation modelsPALs (agentic AI humans) product launchSDK/API for enterprise custom AI employeesUse cases: L&D, healthcare, go-to-marketFounder lessons: conviction, momentum, speed

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome