Partnering with Pace: Making Work Weightless
Sequoia by Bryan Schreier and Lauren Reeder
Pace says its AI agents can handle insurance paperwork inside real companies. That matters because it’s already running in big, regulated insurers, not just demos.
Based on reporting by Sequoia, by Bryan Schreier and Lauren Reeder — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
A Tuesday morning in insurance can still look like a paper chase from another century: a 300-page claims file lands, someone cross-checks clauses, retypes data, and sends pieces off across a pile of systems and vendors. Pace is trying to make that grind disappear. Its pitch is simple enough: forward the submission, step away, and let agents do the reading, the reasoning, the system-hopping, and the draft decisioning in the background.
The company says this is already happening inside some of the largest and most regulated insurers in the world, with tens of thousands of tasks flowing through Pace’s AI agents every month. That’s the real hook here. Plenty of companies can show a clever demo. Far fewer can survive the mess of insurance operations, where accuracy matters, audit trails matter, and permissions matter even more.
Pace’s work sounds less like flashy automation and more like plumbing. The team has to learn standard operating procedures, plug into core systems, meet reliability and accuracy SLAs, and keep a record of every decision. It also has to put guardrails around the whole thing. That is the unglamorous part of agent software, and it’s the part most people would rather skip. Pace seems to be doing the opposite.
Founder Jamie Cuffe and his team are betting that this pattern can spread beyond insurance. Sequoia says the same approach should fit financial services, healthcare, and logistics anywhere long documents, strict rules, and careful judgment slow work down. The firm is leading Pace’s Series A, which is a strong signal that the bet is no longer about what agents might do someday. It’s about whether they can already take over the paperwork without breaking the business.
My take — AI-written commentary, not fact-checked reporting
The useful part of this story is not the word agent; it’s the audit trail. Enterprise AI keeps getting sold as magic, but regulated industries only care if the thing can be trusted after the fact. That’s why the boring stuff—SOPs, permissions, SLAs—is where the actual product lives, and why the demos are mostly just expensive wallpaper.
Read more about this at: Sequoia