The problem
Fraud, lending and AI apps need fresh features at inference time, not stale batch data
What they do
Real-time data and compute platform that feeds fresh context to ML models and agents
Traction Team 30 to ~100 in ~2 years; customers incl. Whatnot, Ramp, Socure, Sunrun
Hiring trend
- Open roles
- 19
- Ops roles
- 2
- Headcount
- ~100
Headcount +170% YoY to ~90-100 (LinkedIn); 17 open roles
Hiring in
- San Francisco
- New York
- London
- Remote (Singapore)
Funding
- Last round
- $50M
- Stage
- Series A
- Date
- May 2025
Led by Felicis
- Hiring+1.519 open roles on the live job board (10–29 roles = +1.5)
- Funding recency+0.5Series A · $50M, May 2025 (17 months ago): 12–24 months = +0.5
- Traction+1.5Strong growth: Team 30 to ~100 in ~2 years; customers incl. Whatnot, Ramp, Socure, Sunrun
- Red flags0None found in the last 24 months
- Total3.51.5 + 0.5 + 1.5 − 0 = 3.5 → Tempo (3.5–4.24)
Distance (Half) is a separate read on how long the lead can last. It doesn't change the score.
Distance
Half · edge: hard to leave
Runs inside fraud and ranking systems at Whatnot, Ramp and Socure; team tripled in two years.
Last checked Oct 25, 2026