Harnessing Modularity for Coding Agents

What you’ll learn and how you can apply it

• Apply modularity fundamentals when designing a new service, splitting a monolith, or judging whether a proposed boundary is a good one
• Learn what complexity actually is, and how to spot it in code you already maintain
• Move from principles to practice: module boundaries, interfaces, tests, the calls that shape the system
• Sustain design over the long term

Course description

Nobody enjoys working on a big ball of mud, including the AI coding agents that are now baked into most development workflows. But they won’t complain; they'll just keep the ball rolling. Software engineering has been through half a century of revolutions, evolutions, and hype cycles, but the core design principles that keep software serving the business over the long haul haven’t moved. What’s new is the stakes: Those same principles now decide whether AI agents amplify productivity or quietly pile more mud onto the ball.

Ultimately, good design decisions are important, but making them stick is the hard part. Vladik Khononov teaches you the fundamentals of system modularity, how complexity cripples both people and agents, and how to harness modular design to get real leverage out of AI-assisted engineering. You’ll walk through real-life examples, weigh design trade-offs in discussion, and sharpen your intuition about when to reach for each principle. You’ll learn to keep slop from creeping back in, preserve modularity over the long run, and keep your agentic assistants productive.