Claim◆Article
The most complex project (Claude Managed Agents) took six months and required a pre-AI-style planning process with PRDs, not rapid prototyping.
Katelyn Lesse describes how building Claude Managed Agents required extensive upfront planning, a PRD, and coordination across teams—much like pre-AI complex projects—and took six months, though that is still faster than the estimated two years pre-AI. ✦ AI generated
Katelyn Lesse · The Pragmatic Engineer · 2026-07-28 · original ↗
There are products you can jump straight to prototyping, but then there are ones where you need to start by architecting it properly. For example, if we build a TypeScript CLI – which is pretty trivial for what needs to be built – we could go straight to prototyping. But with Claude Managed Agents, we needed to first figure out what we are doing. Of course, we did some upfront prototyping for Managed Agents: hacking and spiking things. But prototyping itself was more about understanding the requirements. Our planning process looked more like a typical pre-AI planning process.
Read full article ↗excerpt · fair-use quotation
- ·Took six months end-to-end
- ·Relied on PRD and upfront architecture, not fast prototyping
- ·Prototyping was used only to understand requirements
- ·Simple products (e.g., TypeScript CLI) go straight to prototype
- ·Complex projects (Managed Agents) need full architectural design
- ·Pre-AI process estimated at two years without AI
Around this claim
This moment responds to
supports → Traditional engineering practices like two-pizza teams, PRDs for complex projects, and the challenge of context switching remain unchanged despite AI.Author (narrator) · The Pragmatic Engineerrebuts → A year-long, multi-engineer rewrite of Bun from Zig to Rust was completed in 11 days using 64 parallel AI agents and $165K in tokens.Jarred Sumner · The Pragmatic Engineer