Nuance Labs — Open Roles · Inference Search Partners

Inference Search Partners · Exclusive Placement

Nuance Labs

Face-to-face AI interaction that feels human

$60M Funded Seattle, WA · On-site 25 People Accel · Lightspeed · South Park Commons Founded 2024 2 Roles Open

Two ways to build here.

Both roles are on-site in Seattle, 5 days/week. Both come with relocation support and visa sponsorship. Both report directly into the founding team.

Research
MTS — RL Research
New PhD Grad  ·  14% ISP Commission
Own RL and post-training for Nuance's omni model from 0→1 — the full stack: rollout generation, policy optimization, reward modeling, evaluation, and the systems that connect them, across audio, video, and language in real time.
Base $250K–$350K
Equity Strong
Visa Sponsored
Must Have
  • PhD in ML / CS / RL completing or completed 2024–2026
  • PPO, GRPO, DPO, RLHF — deep internals, not black box
  • PhD research focused on frontier RL algorithms
  • On-site Seattle, 5 days/week
Strong Signal
  • Internship at OpenAI, DeepMind, ByteDance Seed, or Mistral
  • First-author at NeurIPS, ICML, or ICLR
  • verl, OpenRLHF, SGLang, or vLLM experience
  • Multimodal / audio / video RL work
Stack
PyTorch PPO GRPO DPO RLHF verl OpenRLHF vLLM SGLang CUDA
Systems
MTS — Model Optimization & Inference
Experienced (2+ yrs)  ·  15.74% ISP Commission
Own end-to-end inference optimization across the full model stack — LLMs, audio models, and diffusion-based components. KV cache, quantization (INT8/INT4/GPTQ/AWQ), vLLM/SGLang/TensorRT tuning, CUDA kernels. The latency bar is sub-500ms for live audio/visual AI.
Base $250K–$350K
Equity Competitive
Visa H-1B OK
Must Have
  • 2+ years ML engineering in production (not intern-only)
  • Profiling with NSight or torch.profiler — hard dealbreaker
  • Python + Rust or Go
  • On-site Seattle, 5 days/week
Strong Signal
  • CUDA kernel optimization for inference acceleration
  • Quantization: INT8, INT4, GPTQ, AWQ, FP8
  • KV cache eviction, compression, memory-efficient attention
  • Real-time audio/video systems (WebRTC, streaming)
Stack
Python Rust / Go vLLM SGLang TensorRT-LLM CUDA NSight Kubernetes WebRTC

Face-to-face AI that feels human — built from the ground up.

Most conversational AI avatars today are hacks — a face slapped on a speech-to-speech pipeline, stuck in the uncanny valley: emotionless, mechanical, one-turn-at-a-time. Nuance Labs is building something fundamentally different: a full-duplex audiovisual system that can listen, speak, react, interrupt, and respond like a real person.

Current systems take 2–5 seconds to respond. Natural conversation requires sub-500ms. That's a 10× improvement, and it demands rethinking the entire stack — developing foundation models designed for full-duplex from the ground up.

$60M Total Funding
25 People on Team
2024 Founded
<500ms Target Latency
Backed by Accel Lightspeed Ventures South Park Commons

Research pedigree. Production ambition. No busy work.

A small, fast-moving research team with an exceedingly high bar — bringing on only the very best. Every member has massive ownership, deep trust, and the opportunity to shape both the technology and the company from the ground up.

FM
Fangchang Ma
Co-founder & CEO
MIT PhD in Robotics & ML. Previously Research Manager at Apple, with experience at DJI. Published in top AI conferences.
2,400+ citations
EZ
Edward Zhang
Co-founder & CTO
UW PhD in Computer Graphics. Previously Senior Research Scientist at Apple, with experience at Google and Microsoft.
KY
Karren Yang
Founding Research Scientist
MIT PhD. Previously Senior Research Scientist at Apple, Niantic Labs, Meta Reality Labs, Bosch Center for AI, and Adobe Research.
1,800+ citations
CV
Claudia Vanea
Founding Research Scientist
Oxford PhD in AI. Previously founder at South Park Commons. Developed the first AI computational framework in her field — performance surpassing domain experts.
YS
Yaser Sheikh
Advisor
Former VP of Codec Avatars at Meta. Consulting Professor at Carnegie Mellon's Robotics Institute — one of the world's foremost experts on photorealistic avatars.

What makes this worth the move.

Genuinely hard problem
Sub-500ms full-duplex AI — not an incremental improvement. The whole stack had to be rethought from scratch.
🏗
0→1 ownership
25 people. No layers. You own your system end to end with direct influence on technical direction.
🎓
Elite research DNA
PhDs from MIT, UW, Oxford, CMU. Industry experience from Apple Research, Meta, Amazon AGI, Discord.
💰
Strong comp + visa
$250K–$350K base, meaningful equity, relocation support, and visa sponsorship from day one.
📍
In-person, Seattle
Five days/week in the office. A company that genuinely believes proximity accelerates the hardest research.
🚀
Well-capitalized
$60M from Accel, Lightspeed, and South Park Commons — enough runway to do the work right.

The basics done right.

🏥 Health
HSA with ~$2,000 annual company contribution — roughly 2× big tech standard
🏖 Time Off
15 days PTO + public holidays + full week off year-end
🍱 Food
Lunch, drinks, and snacks every workday
🚌 Commuter
Commuter benefits + relocation support for those moving to Seattle