In this video I talk about something a lot of people stress over: do you actually have to remember everything as a software engineer? My honest answer is no—I don’t retain all the information either. There’s just too much to keep in your head: new technologies, languages, frameworks, and then interview prep on top of that. What I do keep is familiarity. Even if I’m rusty, I usually remember “this pattern exists” or “there’s probably a function or library for this,” and that’s the part that matters in real work. I also share how I approach building features: before I write code from scratch, I look for what already exists—internally in the codebase, in similar features, or online. Copying good patterns (and understanding them) is a legit way to learn, and it makes products more cohesive too. Practically, I’m a huge fan of cheat sheets and lightweight docs—especially for command-line setup stuff you’ll forget the second you switch machines. The big exception is interviews. LeetCode-style interviews are different from day-to-day SWE, and memorization matters more because you can’t Google or use ChatGPT. You need fundamentals (syntax, constructors, classes) and you need common algorithms/data structures ready to go—at least to the level where you can write solid pseudocode and translate it quickly.