What it sells
AI workflow platform.
Business model breakdown
Buyer: who owns the budget?
UnverifiedML / AI platform leads and engineering managers at AI-heavy teams (biotech, geospatial, autonomous systems, robotics); case study quote is from a Senior Engineering Manager at Woven by Toyota.
Trigger: what makes them buy now?
Runs fail and restart from scratch, GPUs sit idle, and data must stay inside the company cloud; site pitches durable runs that resume where a step failed and a sovereign control plane that holds references, never payloads.
Workflow: what inputs become what output?
Write plain Python tasks and workflows, each task declaring its own container image and CPU/GPU/memory; Union control plane orchestrates runs on the customer cloud, fans out work (e.g. durable asyncio), resumes failed steps and versions artifacts for reproducibility.
Economics: what it saves and how it charges
Subscription minimum plus metered usage (actions, vCPU, memory, GPU billed to the second).
Distribution: how it reaches customers
Self-serve trial via AWS Marketplace plus enterprise sales; open-source Flyte funnel.
Funding history and coverage
| Round date | Stage | Amount | Investors | Source |
|---|---|---|---|---|
| 2026-02-25 | — | $19M | — | news.google.com |
Profiles summarize public reporting and company websites. Business-model notes are MaxBC's reading of that material, not confirmed facts; unknowns are left blank rather than guessed. Updated 2026-10-12.