Before I was at Meta, I worked at a networking company for 10+ years. It was very different from Meta as a massive consumer web company. On top of the domain differences, the culture is also very different between these 2 companies.
I'm working on ML at Meta, which is something I'm pretty new to. As a senior engineer who wants to make an impact quickly and make a push for staff, how can I effectively make this domain switch to hit the ground running fast?
First, I recommend this in-depth breakdown of how to write great code faster. This applies to every software engineer at any stage in learning a tech stack, but the advice is still very applicable (and even more urgent/important) here.
That being said, here are my thoughts on onboarding into a new tech stack:
Feel extremely comfortable failing and losing face. That's a key trait to learn quickly in a domain you don't know much about. Do that as early as possible.
If there's a question you are too afraid to ask in your first 90 days because "would that make me sound stupid", think about what happens after 12 months. It's likely that you still don't know much about that subject and now you feel horrified about someone figuring that out. That will negatively impact your capability to make an impact. Looking back from that future the only rationale choice is to ask that stupid question and make sure you learn well from the answer even if it's very uncomfortable.
This is based on the observation I had from Meta Bootcampers I mentored. I've seen a staff engineer who was a deep expert in one subject but wasn't comfortable and proactive enough when ramping up to random subjects in the Bootcamp, and almost got fired from the lack of progress in Bootcamp. I've seen a manager (with no prior iOS experience) picked a simple Apple Watch bug, followed the wiki to set up XCode (and other Meta-specific iOS development environments), and fixed the bug in a week.
The best Meta engineers I saw with this trait of fast learning usually had prior experience as early engineers (sometimes CTOs) of successful startups. In hindsight that makes sense. In an early-stage startup environment, if you fail to learn something quickly, build a prototype tonight, and demo it to the investor tomorrow morning because the situation requires that, your company is dead in the water. Losing face is the last thing to worry about.
Lose the pride of "I'm a senior engineer". Be a noob.
Curious to see how this aged. Did you find success after your switch to ML?