An agentic AI lab building the foundation models
for live video agents.
Founded in 2024 by a team of AI PhDs whose last company was acquired by Google. We are training the most human-like real-time human foundation model: agents that move, listen and react naturally, each with a personality of its own. The result is AI you can face, not just read or hear.
Presence and personality, generated live.
Real time, on one GPU
Face, voice and body, live on a single GPU.
Any form, any style
Photoreal, stylized or non-human. All alive.
Distinct personalities
Not one script in many faces. Behavior is what people trust.
Our own data engine
atmee.ai feeds back thousands of hours of real conversations every month.
Bringing AI to life, by giving it
a presence and a mind of its own.
1 image + 15-sec audio
Causeur
The behavior model that brings an agent alive. Personality, emotion, timing and active listening, moment to moment.
Atmee v2
A foundation model for live avatars. Face, body, voice, expression and emotion, generated together in real time.
Live video agent
Causeur decides, Atmee v2 renders. One agent that looks, sounds and behaves like someone.
Serial AI founder and top researchers in computer vision.
Founded October 2024. Six AI PhDs, 100+ publications, 15k+ citations, from Harvard, UMich, Google, Apple, and NEC Labs.
Harvard AI postdoc; AI PhD, Stony Brook. Founder & CEO of Phiar (acquired by Google, 2022). Published in NeurIPS, ICCV, ECCV, PNAS, Psych Science.
AI PhD, UC Merced. Head of AI at Phiar; AI tech lead at Google; researcher at NEC Labs. Area chair at NeurIPS, CVPR, ICCV, ECCV, AAAI, ICLR, ICML.
AI PhD, CEA France. Head of Engineering at Phiar; head of automotive perception at Google. Published in ICCV, BMVC.
Build the next generation of agents with us.
Design partnerships, API access, research collaborations, press, or just a conversation.