About the role
Companies are letting AI agents act on their own: issuing refunds, buying stock, changing prices, moving money. Most can't answer three questions when something goes wrong: which agent did it, was it allowed, and who approved it.
The product is the identity and trust layer for AI agents. Every agent gets its own cryptographic identity and scoped permissions. Actions over a limit wait for a person to approve, a kill switch stops an agent or a whole group of agents in seconds, and every step leaves a signed record that an auditor or regulator can check without trusting us.
It drops into existing stacks through SDKs (npm and PyPI), an MCP connector for ChatGPT, and integrations with agent frameworks, payments and commerce platforms.
WHO WE SELL TO
1. Regulated companies deploying agents. Fintech, payments, insurance and health, where an unexplained agent action is a compliance problem. Buyers: CTO, CISO, Head of AI, Head of Risk.
2. Agent platforms and marketplaces. Companies whose customers' agents act through their APIs and need permissions, approvals and audit for them. Buyers: CTO, Head of Platform, VP Product.
3. Commerce teams running agents. Sellers on Shopify, Amazon, eBay and TikTok Shop whose agents handle refunds, pricing and ads spend. Buyers: Head of Ecommerce, Operations.
Why now: the EU AI Act's high-risk duties, Ofgem's agentic AI guidance and the NCSC's guidance on agentic AI all ask for human oversight, attribution and the ability to shut an agent down. The product gives evidence for each.
WHERE WE ARE
- Product live, with a public playground
- SDKs published on npm and PyPI
- Integration partnership with TruAnon (verified humans behind agent approvals), with a public case study
- Early design partners in fintech and health; pilots in discussion
- Accepted into NVIDIA Inception
- Team: a founder and two co-founders, with backgrounds in GOV.UK, NHS and enterprise delivery
WHAT WE NEED
A senior operator, or two, to turn founder-led conversations into a repeatable pipeline:
- Sharpen our positioning and messaging for the three buyer groups
- Build and run outbound to CTOs, CISOs and Heads of AI
- Turn design partners into paid pilots, and pilots into contracts
- Set up the basics: CRM, pipeline reporting, a short sales deck and a pricing conversation
A GOOD FIRST 90 DAYS
Day 30: Messaging and target list agreed for one buyer group; outbound running; CRM in place
Day 60: 25 qualified first conversations; 5 active pilot discussions
Day 90: 3 signed paid pilots; a pipeline report we can show investors
We'd want milestones and any bonus tied to these outcomes (qualified meetings, signed pilots, revenue), not to activity.