System design interviews often focus on theoretical complexity, but how do Senior Engineers at GitHub actually approach scaling? In this episode, Bassem Dghaidi breaks down how to think about system design when real business impact is on the line.
We discuss why "simple is complicated enough," the dangers of premature scaling, and why vertical scaling often beats complex distributed systems. If you want to bridge the gap between theory and practice, and understand how to design software that actually serves the business, this conversation is for you.
In this episode, we cover:
- The "Order of Magnitude" rule for scaling systems
- Why GitHub often runs millions of requests on simple architecture
- How to communicate technical constraints to non-technical stakeholders
- Why 90% of Bassem's code is now written by AI agents
Connect with Bassem Dghaidi:
https://www.linkedin.com/in/bassemdghaidy
Timestamps:
00:00:00 - Intro
00:00:48 - Theory vs. Practice in System Design
00:02:06 - The Startup That Almost Failed via Kubernetes
00:03:33 - How GitHub Scales (It's Simpler Than You Think)
00:05:20 - The Underrated Power of Vertical Scaling
00:08:23 - Why Big Tech Interviews for Scale You Don't Need Yet
00:10:39 - Software Evolves, It Isn't Just "Built"
00:11:53 - Only Design for the Next Order of Magnitude
00:15:39 - Stop Building Generic Frameworks
00:18:17 - "Hacking" the System Design Interview
00:21:29 - Translating Tech Problems to Business Risks
00:27:37 - Layoffs & Engineering Efficiency
00:29:41 - Proving Your Impact with Numbers
00:31:00 - Professional Engineering vs. Hobby Coding
00:32:19 - "Simple is Complicated Enough"
00:35:03 - The Rise of AI Coding (The Motorcycle Analogy)
00:37:30 - "90% of My Code is Written by AI Agents"
00:41:04 - How to Become a Great Engineer
#SystemDesign #SoftwareEngineering #GitHub