# Nuance Labs: Seattle AI Profile — Product, Footprint & Verified Record

- **Canonical Web Article:** https://tygartmedia.com/nuance-labs-seattle-wa-ai-profile/
- **Profile Brief (PDF):** https://tygartmedia.com/wp-content/uploads/2026/11/nuance-labs-seattle-wa-ai-profile.pdf
- **Markdown Twin:** https://tygartmedia.com/wp-content/uploads/2026/11/nuance-labs-seattle-wa-ai-profile.md
- **Snapshot Infographic:** https://tygartmedia.com/wp-content/uploads/2026/11/nuance-labs-seattle-wa-ai-profile-infographic.png
- **Verified Date:** 2026-11-19

> **Sourced Baseline:** As of November 19, 2026, Nuance Labs operates independently from Seattle, Washington, with a live website at nuancelabs.ai showcasing interactive audio-visual AI models and developer API access. Backed by $60M in disclosed venture funding through its September 2026 Series A, the research lab employed 24 to 27 people across Fall 2026 GeekWire reporting and is unrelated to Microsoft-owned Nuance Communications.

## Verified Company Record

| Field | Verified Record (As of Nov 19, 2026) |
| :--- | :--- |
| **Company** | Nuance Labs |
| **Official Website** | https://www.nuancelabs.ai/ |
| **Washington HQ** | Seattle, WA |
| **Other WA Footprint** | Early team operated out of Foundations in Seattle (Sep 2025 GeekWire); Co-Founder & CTO Edward Zhang holds a UW PhD |
| **Founded** | 2025 (early 2025, per GeekWire) |
| **Founders / Leadership** | Fangchang Ma (Co-Founder & CEO), Edward Zhang (Co-Founder & CTO), Karren Yang (Co-Founder & Chief Scientist) |
| **Stage** | Series A |
| **Announced Funding** | $60M total ($10M Seed in Sep 2025 + $50M Series A in Sep 2026) |
| **Team Size** | 24 (Fall 2026 GeekWire 200) to 27 (Sep 15, 2026 GeekWire report) |
| **Primary Ecosystem Ties** | GeekWire 200 (Fall 2026, No. 163), Foundations Seattle, University of Washington PhD founder lineage, South Park Commons |

## What Nuance Labs Builds

Nuance Labs is a Seattle foundation model research lab developing a unified full-duplex audio-visual model designed to perceive and generate real-time human expression simultaneously. Rather than chaining separate speech transcription, text generation, voice synthesis, and facial animation tools in sequential turns, the company is training a single system that sees, hears, reasons, speaks, and expresses concurrently.

According to [Nuance Labs' official site](https://www.nuancelabs.ai/) and [September 2026 GeekWire coverage](https://www.geekwire.com/2026/seattles-nuance-labs-raises-50m-to-give-ai-models-human-expression-and-nuance/), its architecture applies auto-regressive transformers to predict the next audio and visual tokens frame by frame. The system models conversational cues beyond literal words—such as tone shifts, gaze, brief pauses, nodding, interjections, and backchanneling—while treating 500 milliseconds as a hard ceiling for conversational latency.

- Unified full-duplex engine processes incoming audiovisual signals and streams synchronized facial and vocal responses without turn-taking stalls.
- Specialized Human Foundation Model predicts next audio and visual tokens directly instead of relying solely on text-first language pipelines.
- Designed around a 500ms conversational latency target and active-listening behaviors including nodding, pauses, and vocal backchanneling.
- Targets enterprise and developer applications where real-time human expression affects outcomes, including sales, customer service, professional coaching, and education.

## Washington Footprint and Leadership

Nuance Labs lists Seattle, Washington as its headquarters on its live website and anchored its operations locally after early team members worked out of Foundations, a Seattle startup community. In a [September 2025 GeekWire profile](https://www.geekwire.com/2025/ex-apple-phds-raise-10m-choose-seattle-to-build-an-ai-foundation-model-that-understands-emotion/), co-founders Fangchang Ma and Edward Zhang explained that they initially considered relocating to San Francisco during fundraising before choosing to stay in Seattle to recruit from local engineering networks and build a dedicated local research lab, while noting plans to establish a Bay Area footprint later as the company scales.

The three co-founders are former Apple PhD researchers who bring complementary backgrounds in robotics, graphics, and audiovisual synthesis. Readers researching the company should also note two common name distinctions: Seattle's Nuance Labs (nuancelabs.ai) is completely independent of legacy speech software company Nuance Communications (acquired by Microsoft in 2022) and unrelated to Nuance Labs Consulting, a separate marketing and design consultancy.

- Fangchang Ma (Co-Founder & CEO): Holds a PhD in robotics and machine learning from MIT; joined Apple in 2019 and previously served as an Engineering Manager at Apple's Seattle office.
- Edward Zhang (Co-Founder & CTO): Holds a PhD in computer graphics from the University of Washington; previously served as a Senior Research Scientist at Apple in Seattle, where he met Ma.
- Karren Yang (Co-Founder & Chief Scientist): Holds a PhD in audio-visual synthesis from MIT; previously served as a Senior Research Scientist at Apple.
- Broader research bench: Nuance Labs states on nuancelabs.ai that its staff includes PhDs from MIT, UW, Oxford, CMU, and Johns Hopkins with prior experience at Apple, Meta, and Amazon AGI.

## Products, Users, and Public Traction

Nuance Labs presents two primary public surfaces on [nuancelabs.ai](https://www.nuancelabs.ai/): interactive face-to-face AI experiences (including an interactive persona presented as Emma) and developer API access. Accompanying its September 2026 Series A announcement, the company released a demonstration video illustrating an active-listening avatar adjusting facial expressions and verbal timing dynamically as a person speaks.

Per GeekWire's September 15, 2026 report, Nuance Labs planned to open a public research preview later in 2026 so outside users could test its face-to-face interactive avatar firsthand. Named enterprise customers, commercial pricing tiers, and revenue figures are not publicly disclosed.

- Public product surfaces: Interactive AI demonstrations ('Meet Our AIs' / 'Meet Emma') and a developer API waitlist/access portal ('Get the API').
- Planned milestone: Public research preview of its interactive audiovisual avatar announced for later in 2026 (Sep 15, 2026 GeekWire).
- Hiring focus: Deploying Series A capital to recruit researchers and engineers across modeling, data, evaluation, and real-time serving.
- Named customers and revenue: Not publicly disclosed as of November 19, 2026.

## Funding, Stage, and Place in Washington's AI Landscape

Nuance Labs has raised $60M in total disclosed venture capital across two institutional rounds within roughly a year of its early 2025 founding. The company first disclosed a $10M Seed round led by Accel in September 2025—with participation from Lightspeed Venture Partners and South Park Commons—when it had four employees. One year later, on September 15, 2026, Lightspeed led a $50M Series A joined by Accel, South Park Commons, NVIDIA, and Define Ventures.

Headcount reporting reflects rapid expansion following the seed round, with two slightly different Fall 2026 public figures: [GeekWire's September 15, 2026 funding story](https://www.geekwire.com/2026/seattles-nuance-labs-raises-50m-to-give-ai-models-human-expression-and-nuance/) reported 27 employees, while the [Fall 2026 GeekWire 200](https://www.geekwire.com/geekwire-200/) snapshot recorded 24 employees (ranking the company No. 163, up 20 spots). Within Washington's AI landscape, Nuance Labs stands out as a Seattle-headquartered foundation model lab funded primarily by national institutional and strategic investors rather than relying exclusively on local seed funds.

- Seed round ($10M, Sep 2025): Led by Accel with Lightspeed Venture Partners and South Park Commons.
- Series A ($50M, Sep 15, 2026): Led by Lightspeed Venture Partners with NVIDIA, Define Ventures, Accel, and South Park Commons, bringing total disclosed funding to $60M.
- Team growth: Expanded from 4 employees at its September 2025 seed announcement to 24 (Fall 2026 GeekWire 200) or 27 (Sep 15, 2026 GeekWire).
- Regional index presence: Ranked No. 163 on the Fall 2026 GeekWire 200 index of privately held Pacific Northwest tech startups.

## Related Washington AI Profiles & Tygart Media Context

- [Hearvana — Seattle On-Device Acoustic & Hearing AI Profile](https://tygartmedia.com/hearvana-seattle-wa-ai-profile/)
- [Ethosphere — Seattle Frontline Retail Voice AI Profile](https://tygartmedia.com/ethosphere-seattle-wa-ai-profile/)
- [Prom.dev — Seattle Consumer Social Group-Chat AI Profile](https://tygartmedia.com/prom-dev-seattle-wa-ai-profile/)
- [Is Your Voice Assistant a Brain, or Just a Router?](https://tygartmedia.com/is-your-voice-assistant-a-brain-or-just-a-router/)

## Verified Primary Sources

- [Nuance Labs Official Website](https://www.nuancelabs.ai/)
- [Nuance Labs About & Team](https://www.nuancelabs.ai/about)
- [GeekWire — Nuance Labs $50M Series A (Sep 15, 2026)](https://www.geekwire.com/2026/seattles-nuance-labs-raises-50m-to-give-ai-models-human-expression-and-nuance/)
- [GeekWire — Nuance Labs $10M Seed & Seattle HQ (Sep 9, 2025)](https://www.geekwire.com/2025/ex-apple-phds-raise-10m-choose-seattle-to-build-an-ai-foundation-model-that-understands-emotion/)
- [GeekWire 200 Index](https://www.geekwire.com/geekwire-200/)
- [GeekWire — Fall 2026 GeekWire 200 Update (Sep 30, 2026)](https://www.geekwire.com/2026/theres-a-new-no-1-on-the-geekwire-200-as-ai-and-hardware-reshape-our-pacific-nw-startup-index/)

## Frequently Asked Questions

### Is Seattle's Nuance Labs affiliated with Microsoft-owned Nuance Communications?

No. Seattle-based Nuance Labs (nuancelabs.ai) is an independent Series A AI research startup founded in early 2025 by former Apple PhD researchers. It has no affiliation with legacy speech company Nuance Communications, which Microsoft acquired in 2022, or with Nuance Labs Consulting.

### How much funding has Nuance Labs raised, and who backed the company?

Nuance Labs has disclosed $60M in total venture funding across two rounds: a $10M Seed round led by Accel in September 2025 and a $50M Series A led by Lightspeed Venture Partners in September 2026. Other disclosed backers include NVIDIA, Define Ventures, and South Park Commons.

### How does Nuance Labs' full-duplex model differ from conventional voice and avatar pipelines?

Conventional voice avatars typically chain separate models for speech-to-text, text generation, voice synthesis, and facial animation in sequential turns. Nuance Labs is training a single auto-regressive audio-visual foundation model that predicts next audio and visual tokens simultaneously, targeting conversational latency below 500 milliseconds with real-time nodding, gaze, and tone shifts.
