What it sells
AI agent security.
Business model breakdown
Buyer: who owns the budget?
UnverifiedHead of AI or engineering lead at companies running production agents (hypothesis from tiering and CLI/MCP developer tooling).
Trigger: what makes them buy now?
UnverifiedFrontier-model cost and agent failures in production; site argues you do not need a frontier model for every task and offers to measure small models against the current one.
Workflow: what inputs become what output?
Connect repo and production traces to find failing steps; build training datasets from traces in Data Workshop; evaluate prompts and models against real tasks; fine-tune an open model and deploy it behind an OpenAI-compatible API (weights downloadable).
Economics: what it saves and how it charges
Subscription tiers (Free, Pro $120/month, Enterprise) plus usage credits consumed by training, inference, optimizer runs and Data Workshop chat; $50 one-time starter credit.
Distribution: how it reaches customers
Self-serve console with CLI and MCP server, plus demo-led enterprise.
Funding history and coverage
| Round date | Stage | Amount | Investors | Source |
|---|---|---|---|---|
| 2026-02-11 | — | €2.3M | — | 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 2026-10-13.