A book in progress · by Keshav Suman
The Field Guide to Autonomous Agents.
A practical handbook for building agents that use tools, recover from failure, and finish real work.
A forthcoming book about the systems, product decisions, and safety boundaries behind useful autonomous agents. It distils the field notes here into a practical path from tool contracts and memory to verification, evaluation, and human approval.
The book is currently in progress. Join the list for chapter previews and release news.
What you’ll learn
Useful autonomy, end to end.
- Design agents around explicit outcomes and evidence.
- Build reliable tool use, retries, memory, and handoffs.
- Evaluate task completion, not just model responses.
- Place human approval where risk actually demands it.
Subscribe to the blog
Get new posts in your inbox. No spam, unsubscribe anytime.