The problem
Agents can't learn real jobs without realistic, gradable practice environments
What they do
Builds simulated work environments where frontier labs train and test AI agents
Traction Run-rate revenue $1M to $63M in ~6 months (company claim, Jun 2026)
Hiring trend
- Open roles
- ~13
- Ops roles
- 1
- Headcount
- ~100
~100 staff (LinkedIn 103; Harmonic); ~13 roles on own careers site
Hiring in
- San Francisco
- New York
- Remote (US)
Funding
- Last round
- $45M
- Stage
- Series A
- Date
- Apr 2026
Led by Sequoia, Menlo Ventures, Bain Capital Ventures (insider round)
- Hiring+1.513 open roles on the live job board (10–29 roles = +1.5)
- Funding recency+1.5Series A · $45M, Apr 2026 (6 months ago): 6 months or less = +1.5
- Traction0No growth signal counted: Run-rate revenue $1M to $63M in ~6 months (company claim, Jun 2026)
- Red flags0Watch-outs noted below, none scored as a significant flag
- Total3.01.5 + 1.5 + 0 − 0 = 3.0 → Steady (1.5–3.49)
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
Revenue jumped to $63M run rate, but its lab buyers could build their own training gyms.
Watch-outs
Aug 2026: offboarded ~80% of generalist contractors without notice; separate group reports unpaid June work
Sources: AI Trainer Work, Aug 2026 ↗
Open ops roles · 1
Last checked Oct 13, 2026