It seems pretty well-accepted that AI coding tools struggle with real production codebases. At AI Engineer 2025 in June, The Stanford study on AI's impact on developer productivity found:

1. A lot of the ""extra code"" shipped by AI tools ends up just reworking the slop that was shipped last week.

2. Coding agents are great for new projects or small changes, but in large established codebases, they can often make developers less productive.

The common response is somewhere between the pessimist ""this will never work"" and the more measured ""maybe someday when there are smarter models.""

After several months of tinkering, we've found that you can get really far with today's models if you embrace core context engineering principles.

This isn't another ""10x your productivity"" pitch. I tend to be pretty measured when it comes to interfacing with the ai hype machine. But we've stumbled into workflows that leave me with considerable optimism for what's possible. We've gotten claude code to handle 300k LOC Rust codebases, ship a week's worth of work in a day, and maintain code quality that passes expert review. We use a family of techniques I call ""frequent intentional compaction"" - deliberately structuring how you feed context to the AI throughout the development process.

In this talk, I'll share what we've learned since first sharing these techniques back in August, and some educated predictions on what's coming in the next 6-12 months for software engineers.

Speaker: https://twitter.com/dexhorthy

AIE is coming to London and SF! see dates and sign up to be notified of sponsorships, CFPs, and tickets: https://ai.engineer

AI Engineer

A deep dive into advanced context engineering techniques for AI coding agents, focusing on the research-plan-implement (RPI) workflow to solve complex problems in brownfield codebases. The talk introduces concepts like intentional compaction, the "dumb zone" (context window degradation beyond 40%), and sub-agents for context control. Key insights include treating context windows as the primary optimization target, avoiding semantic diffusion in terminology, and maintaining mental alignment through readable plans. The speaker demonstrates successful application in large Rust codebases and emphasizes that AI amplifies thinking rather than replacing it, requiring human oversight at high-leverage points in the workflow.

No Vibes Allowed: Solving Hard Problems in Complex Codebases – Dex Horthy, HumanLayer