Life update: started a job as a creator-in-residence at @clay_gtm. I get to build, research, and storytell for a living😀 For my 1st YT video, I went down a silly rabbithole that turned quite meaningful. I've never met a Bhaumik in my life, I thought Clay could fix that

In 13 minutes, @jeffbarg, Vyshu Khota, and Soroush Khadem walk through how Clay scaled agent evals agents at 300M+ runs a month. Topics covered: ✅ Their four quadrant eval framework ✅ Why closing the production-to-eval loop is the hardest part ✅ How a data lake and long context changed what agents can do with data

And just like that, the question was no longer how to keep the agents in the sandbox. It was whether intelligence, like love, could ever survive inside something built to contain it.

Not to call the game early, folks. But checkmate. Account Agents are here. You can now add an Account Agent to a Clay Workflow. It remembers what it concluded last time and chooses the next step from a list of configurable actions. Then Workflows, our visual canvas for triggers and actions, carries it out: branching the play, running enrichments, notifying you in Slack, or updating records. The result gets saved back to your account data automatically. In technical terms: Build Workflows through the CLI, store everything in Audiences, and run Account Agents on top of it. Account Agents in Workflows are now in open beta for Enterprise, Growth, and Launch customers. Read how they work → What does this have to do with chess? Mishti breaks it down 👇

I know it's quite gauche to admit you still use plan mode, but-- Codex plan mode is *so* good. The iterative Q&A is exactly how I'd want to talk to a smart thought partner when I'm thinking through an idea.

say it with me now. experts are fake, smart generalists rule the world, everything is designed by people no smarter than you, and courage is in shorter supply than genius. now, prove the Riemann hypothesis. enough with the partial results.