AI-Native

How I Built a Software Factory with AI Agents

31 days. 31 products. One AI agent. A hands-on journal of building an AI-native software factory from scratch β€” with real code, real failures, and real timestamps.

Author

Venkatesh Rao

Published

January 1, 2026

31 Days. 31 Products. One AI Agent.

πŸ“ Current Status: February 23, 2026 β€” Pre-Sprint Setup Phase

This book is a real-time record of my attempt to build a software company using AI agents as the primary engineering team.

The Plan: From March 1 to March 31, 2026, build one product every day. My only engineering resource is an AI agent named Jarvis β€” Claude running on OpenClaw, talking to me through Telegram, with access to my servers, my code, and my deployment pipelines.

What you’re reading: A work-in-progress book that gets updated daily as the work happens. No retrospective storytelling. No post-hoc analysis. Just real events documented as they occur.

Part I covers the setup period (February 2026) β€” getting the agent running, solving the first real problems, and preparing for the sprint.
Part II will cover each of the 31 days (starting March 1).
Part III will capture the patterns and lessons (written as they emerge).

Who This Book Is For

  • Developers who want to understand how to work with AI agents on real projects
  • Founders who want to see what β€œAI-native” means in practice, not in theory
  • Anyone who wants an honest account of what works and what does not

How to Read This Book (Work in Progress)

Part I (available now) should be read in order. It explains the tools, the setup, and the context for the 31-day sprint.

Part II (starting March 1) will be written daily as each product is built. Each day will be a self-contained story.

Part III (written as patterns emerge) will capture lessons and insights as they become clear.

TipFollowing the Progress

Check out different git commits to read the book as it existed on any given day. End-of-day commits will be tagged for easy access.

The code is open source. Every repository is linked in the relevant chapter. You can read the code, fork it, or use it as a starting point for your own projects.

TipSource Code

All code lives at github.com/venkatesh3007. Each chapter links to the relevant repositories and commits.