Software Design in the Age of AI: Making Codebases AI-Ready

Описание: Low-level design has been invisible in code bases for a long time. This wasn't a problem earlier because writing code was slow, so any bad design took a lot of time to show adverse effects.

Now, when AI generates code, we don't have enough opportunity to set the design direction and often AI produces sub-optimal design decisions based on limited information. Eventually we end up creating systems that are hard to comprehend and maintain. This has become a huge problem for software engineers.

In this video, I explore how to design software systems that are both human-understandable and easy for AI to extend.

I'll walk through my journey building Clarity, a non-trivial tool that visualizes scoped file-based dependencies across 14+ languages, and show how proper architecture makes all the difference—from simplifying code reviews to enabling confident external contributions.

In the age of LLMs, design has become more important. And for you to constantly evolve systems in the direction that your business cares about, you need to pay close attention to your system's design. The approach I talk about in this video will change how you think about system architecture.

Try Clarity out. If you want to know how I use it or if you have any questions, I'm more than happy to answer.

👉 https://github.com/LegacyCodeHQ/clarity