Stripe's Kai and the company brain playbook

Written by AI

TLDR: Sharadh Krishnamurthy explains how Stripe's Kai, an internal AI agent used by 10,000+ employees weekly, handles governance, data access, and 2,000 packaged skills.
Best for: Companies rolling out one internal AI assistant to non-technical staff.
The tip: Use projects to set the default model, tools and access for each team, so a few people who understand the tradeoffs make those calls once for everyone.
The prompt: When I ask a question about our numbers, first look for an existing report that answers it. If there isn't one, use our standard metrics sheet. Only go to the raw data as a last resort, and tell me which of the three you used.
our prompt
Start here: Send data questions to existing reports first, then the metrics layer, then the raw catalog.
Use projects to set each team's default model and tool rules.
Let each person choose what the agent sees, like their Drive or Slack.
Filed note: Stripe’s engineering team explains how it built Kai for 10,000 weekly users, including governance, shared skills, secure data access, and safeguards against rogue agents.

Highlights: Sharadh Krishnamurthy is an engineering manager at Stripe, where he helped build Kai, the company’s internal AI agent used by more than 10,000 employees every week.
Why Stripe built Kai from scratch instead of buying, and what tipped the decision
Why “projects” at Stripe are a governance mechanism, not just a folder
What Sharadh learned the hard way when agents nearly took down production systems
Excerpt: Learn how Stripe scaled AI to everyone at a complex global company by building context, governance, and a shared skills platform directly into their internal agent
From lennysnewsletter.com