TOP SKILLS

TOP SKILLS tech companies are ACTUALLY hiring for Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Mon Tues Wed Thurs Fri Learn how we count contributions Less More [From a new grad SWE at meta] PROBLEM DECOMPOSITION the new bar for junior engineers is taking a vague, half-formed problem and turning it into a scoped, executable plan. AI handles the boilerplate. your job is figuring out what to build and how to break it down - without someone holding your hand. think the reasoning skills behind writing a really good RFC. To RFC or not! START Is there more than one solution? NO Do you have to pick a new library service? YES! KINDIN'? YOU KNOW IT'S ITS RFC TIME. STRAIGHT TO CO-DOWN RIGHT AWAY. OWNERSHIP MENTALITY the engineers who get promoted aren't just the ones who execute well. they're the ones who proactively flag risks before their manager sees them, ask the clarifying question before the project goes sideways, and treat their work like it's theirs - not like they're waiting for the next task. especially now with AI handling execution, the things that actually distinguish you are the judgement calls - knowing when something is wrong, raising it before it becomes a problem, and caring enough to follow through. nobody is coming to manage your career for you. @mareomni SYSTEM DESIGN INTUITION what separates you is walking into a design review and justifying your tradeoffs before anyone asks. senior engineers are watching whether you think in systems, not just in code. can you articulate why you chose this architecture over that one? can you identify where your design breaks under scale before someone else does? the person who can do that as a new grad stands out immediately - because most can't. @mareomni DEBUGGING (AI-GENERATED CODE) vibe coding gets you 80% there and then leaves you with a bug that's subtle, contextual, and completely non- obvious. and you can't just ask AI to fix it - when it wrote the bug, it doesn't have the system context to diagnose it objectively. it'll go in circles or double down on the wrong assumption. the engineers who actually understand debugging principles - who can trace through unfamiliar code, identify what assumption broke, and fix it without breaking the rest of the system - are very valuable. that skill doesn't get outsourced. @mareomni CODE REVIEW LITERACY AI can write code but it cannot always review it objectively. the human value- add has shifted from writing to validating - knowing when an abstraction is going to cause pain later even if it works today, spotting edge cases that are missed by the prompting, and recognizing when something technically runs but doesn't belong in your codebase. @mareomni CROSS-FUNCTIONAL COMMUNICATION you may work with PMs and designers who do not care how elegant your solution is. they care whether it ships and whether you flagged the risk early. the engineers who get recognized aren't always the most technically brilliant - they're the ones who can translate a complex technical decision into a sentence that makes sense to a non-engineer, without dumbing it down or over- explaining. that skill is rarer than you think, and it compounds fast. @mareomni Follow @mareomni For more software engineering advice and tips