Newsroom

When Agentic AI Scales, the Provenance Gap Scales With It

Pace, which describes itself as an AI operations partner for the world's largest insurers, has closed a $46 million Series B led by Thrive Capital and Sequoia Capital.

Newsroom 27 May 2026

Pace, which describes itself as an AI operations partner for the world's largest insurers, has closed a $46 million Series B led by Thrive Capital and Sequoia Capital. The company says it will use the funding to scale agentic AI workflows to tens of millions of insurance operations tasks this year. Customers include WTW, Prudential, and Convex.

The investment is a clear signal. Agentic AI systems that don't just assist with decisions but autonomously execute them, is moving from proof of concept to production infrastructure in insurance at pace. Policy servicing, claims handling, data ingestion for new business and renewals: these are no longer tasks being reviewed by AI. They are being completed by it.

This is a significant development, and not only because of what it enables. Scale of this kind introduces a structural question that most AI operations platforms are not yet equipped to answer: when an agent makes a consequential decision autonomously, can you reconstruct exactly what it knew, what it weighted, and why it reached that conclusion?

A quarter of a million completed insurance workflows, as Pace reports, is a quarter of a million decisions. Each one sits inside a regulatory environment that is tightening its requirements around traceability, explainability, and audit. The EU AI Act's high-risk provisions apply directly to AI systems used in essential services and financial decision-making. The agents doing this work will need to prove their reasoning, not just record their outputs.

Agentic AI at scale is an impressive capability. Provable agentic AI at scale is the harder and more valuable problem. That is the gap the market is moving towards, and it is the gap Panamorphix was built to close.