.jpg)
The problem
Voice AI teams can't get enough natural, speaker-separated conversation audio to train and test realistic speech models
What they do
Designs, records and licenses conversational audio datasets and evals for voice AI models
Traction Eight-figure revenue run rate within a year (May 2025); ~$43M revenue est. (Tracxn); most Mag7 and leading audio labs as customers
Hiring trend
- Open roles
- 17
- Ops roles
- 4
- Headcount
- ~50 FTE (LinkedIn ~120-130 incl. contractors)
~15 → ~50 staff in a few months (Jul 2026); job postings +67% in 2026 (Revelio)
Hiring in
- San Francisco
- New York
- Remote (US)
Latest news
Oct 2026
Released DAI-ASR-I18N, a 21-language conversational speech-recognition benchmark ↗Glassdoor
3.6 ★ 2 reviews ↗Funding
- Last round
- $50M
- Stage
- Series B
- Date
- Oct 2025
Led by Meritech (with NVIDIA); Series A led by Alt Capital, Amplify
- Hiring+1.517 open roles on the live job board (10–29 roles = +1.5)
- Funding recency+1Series B · $50M, Oct 2025 (12 months ago): 6–12 months = +1
- Traction—Eight-figure revenue run rate within a year (May 2025); ~$43M revenue est. (Tracxn); most Mag7 and leading audio labs as customers
- Red flags—None on file
- Total3.5Stored score → Tempo (3.5–4.24)
Hiring and funding from the stored fields add up to +2.5. The stored score of 3.5 includes a judgment call the fields don't capture (often a boundary month or a round we don't count), so traction and red-flag points can't be split out exactly.
Distance (10K) is a separate read on how long the lead can last. It doesn't change the score.
Distance
10K · edge: learns from use
Fast-rising specialist in conversational audio data with a research feedback loop, but sells to a small set of very large buyers who could build this data in-house, and faces Scale AI, Defined.ai, Protege and Panels.
Sources: Contrary Research: David AI deep dive (Sep 2026) ↗ · Bloomberg Law: David AI notches $500M value (Oct 2025) ↗
Last checked Oct 25, 2026