10+ hours of podcasts, in 5 minutes.
7 quotes from 1 episode on How I AI, each with a timestamped link to the source.
Sharadh Krishnamurthy argues that internal AI agents require strict project-level boundaries to operate safely within a company. Department leaders govern these tools by setting default models for simple tasks and requiring human approval before sensitive actions occur.
Autonomous agent loops act as unintended stress tests that can easily crash internal systems.
“Agents are very creative at bringing your infra down…”
Teams handling private information enforce rules that stop accidental data leaks to public documents without banning AI tools entirely.
“Let's say you're a person on the HR team who's dealing with a bunch of sensitive information. You really don't want the agent to sort of go rogue and put that sensitive data into some public Google document that all Stripes can access, but you also don't want to tell them, 'Oh, you can't use any tools because your workloads are too sensitive.'”
From How Stripe Uses Project-Level Controls to Govern AI Agents
Department leaders act as policy makers by treating specific projects as governance folders. They restrict expensive models for basic tasks and flag certain tools as sensitive based on the workload.
“The two things that we were very intentional about is the idea of projects…”
From How Stripe Uses Project-Level Controls to Govern AI Agents
“A project can say, 'Hey, here's the default model we want people to use. We don't even want to let them use these super expensive models because the job that you're trying to do here doesn't need one of these super models to look at them.'”
From How Stripe Uses Project-Level Controls to Govern AI Agents
“Someone who is the DRI of a space can decide that certain tools are sensitive for the workloads that these people are going to be using…”
From How Stripe Uses Project-Level Controls to Govern AI Agents
Systems originally designed to guide human workers provide the foundation for automated tasks. Bots struggle to identify the correct tables or queries without curated platforms and clean data catalogs.
“These investments were made for humans but have held up really well for agents because turns out reasoning through it, agents have the same problem. They can answer the question, but they have no idea if it was the right query or the right table.”
“It took us like one and a half people over two weeks to get Vzero out the door…”
From Why Stripe Built an Internal AI Agent Instead of Buying One
How we attribute quotes. Every quote was matched against the episode transcript, so the words and the timestamp are real (we trim filler words like "um", nothing else). The name comes from our written summary of the episode. YouTube gives us no voice-by-voice transcript, so open the timestamp to hear who is talking. See a wrong name? Tell us and we fix or remove it.
10+ hours of podcasts, distilled into one 5-minute read. Free, every Sunday morning.
One email a week. Unsubscribe with one click.