Skip to content

Blog

Onboard your AI assistant like a junior developer

Teams spend weeks onboarding a junior developer and minutes onboarding their AI assistant. Both need the same thing: context, standards, and patience.

Written by

Context engineering

Published: 6 November 2025 Updated: 2 min read

Your team spends 8 weeks onboarding junior developers.

Your AI assistant? 30 seconds and a prayer.

Last month, an engineering lead told me: “We spent 12 weeks onboarding our junior developer. We gave our AI assistant 10 minutes and expected magic.”

This is why most teams are frustrated with AI coding tools.

AI assistants went from helpful autocomplete to autonomous contributors. The old way doesn’t scale when AI writes a large share of your codebase.

AI assistants need structured onboarding, just like junior developers.

What both need on day 1

A junior developer on day 1 asks: Where’s the repo? What’s our coding style? Who reviews my PRs?

An AI assistant on day 1 needs to know: What’s your architecture? What patterns do you follow? How do you validate output?

Both need the same thing: context, standards, and patience.

The 8-week onboarding framework

Weeks 1 to 2, foundation building:

  • Share architectural documentation
  • Start with isolated, simple tasks
  • Establish feedback loops and document patterns

Weeks 3 to 4, skill development:

  • Test understanding through progressive complexity
  • Put validation processes in place
  • Build trust and define quality boundaries

Weeks 5 to 8, production ready:

  • Delegate complex, multi-file features
  • Focus on requirements over implementation
  • Scale successful patterns team-wide

What changes with structured onboarding

Teams without structured onboarding:

  • Most AI suggestions rejected in code review
  • Engineers spend more time fixing than shipping
  • “AI doesn’t understand our codebase” becomes the default excuse

Teams with structured onboarding:

  • Most AI code passes review the first time
  • Long tasks complete reliably and in a fraction of the time
  • Engineers advocate for AI adoption across the organization

The same engineering practice as for every team member

This isn’t about treating AI like a human. It’s about applying the same engineering practice we use for every team member.

Junior developers become senior contributors because you invest in their understanding of your systems, patterns, and practices. Your AI assistant can do the same, but in a structured way, not by magic.

Your AI assistant wants to be a good team member. Give it the same structured onboarding you’d give any junior developer.

Have you onboarded your AI assistant like you would a junior developer?

Further reading

  1. Agents fail from amnesia: context as a living playbook Agent failures are context failures. The ACE framework treats context as a playbook that grows through small updates instead of rewrites, and small open models gain the most from it. 2 min readShort post
  2. Reduce and delegate: two ways to keep an agent's context focused A context window is not infinite RAM. First cut what never belongs in it, then spread the rest across sub-agents and context bundles so long tasks keep their focus. 3 min readShort post
  3. Who owns the file every agent reads first Since Claude Code 2.1.277 reads AGENTS.md where no CLAUDE.md exists, one instruction file can serve every coding agent on a team. The file was never the hard part: the instructions are a team decision and need one owner, so merge the files into one AGENTS.md and name who owns it. 1 min readShort post

All 16 posts on Context engineering

ClientsCase study: A large DIY retailer in Germany

Test the constraint in your system

The Factory Readiness Assessment connects this research to your value stream, your data, and the next decision.

Training on this topic: Agentic Engineering for Developers and QA
See the assessment

Want to check one team first? The self-check: twelve yes or no questions, three minutes, instant result, no account. Take the self-check