Stories from building with AI agents without losing the plot.

No CMS. No database. Just Markdown in git, written like field notes: the problem, the bad first try, the fix, and the happy path. If the story feels like your team, there is a clear next step to work with Sandip or join the newsletter.

I help teams turn agent experiments into useful systems.

If your team is trying to build with AI agents, I can help you shape the workflow, write the specs, review the architecture, or build the first production-ready agent with you.

Where I can help

  • Agent workflow design for real products
  • Context files, specs, review loops, and verification
  • Hands-on builds for internal tools, content systems, and product agents

What should come next

How to design an AI agent discovery call for a client project
The beginner-friendly repo template for AI agent projects
Why one task per agent run beats giant prompts
How to review AI-generated pull requests without reading every line twice
The newsletter notes behind a production agent workflow
How to recover when an AI agent deletes the wrong thing
A simple verification checklist for static Astro projects
How to turn agent mistakes into reusable consulting offers