After 10 years as a full stack SWE and eng manager, I pivoted into AI while working at Shopify, and was recently laid off as Head of AI at a collapsing pharma startup. Title is nice, but I only really have two years experience in ML. While it is high quality experience (training and shipping models and LLM apps at web scale)-- I'm feeling a bit scared. I don't have a ton of savings and two kids so I need something soon.
I'm deeply passionate about language models, for the first time in my career working with a particular technology has felt like a real calling-- staying up nights and weekends just to learn and build. My first research paper ever was published at NeurIPS last year.
However, I'm feeling fairly unconvincing as an ML engineer after the layoff. Probably the perfect role would be something in between web and ML. So now we're at the question:
Given that I'm pretty desperate to land anything (3 mos runway before pulling out of investments, wife really against this) I'm wondering how to approach my search:
Some confounders:
I have referrals at Google and Microsoft, but don't want to burn them on ML roles if I'm obviously unqualified having only 2 years ML. I know I can absolutely add value wherever I land, but these feel like precious gold to me, and I don't want to get tossed out of the running for playing it silly. I can likely get some at Meta as well, but again, I don't want to play myself going for stuff that's just inappropriate. This has never been an issue in the past, I've been able to land stretch roles or at least get the interview but stakes are different now and my confidence is lower.
I am a good eng manager, and would do it again, but I have a feeling it's an altogether different search. Is there a way to increase the surface area of possible roles by applying to manager jobs too-- without splitting my energy?
Anyway, its helpful just to think out loud, would appreciate any advice here. Current plan is to create 3 resumes, start blasting applications and networking to get the interview funnel spun up before the leetcode grind.
Thanks.
Update on this, I blasted out 150 apps and got a handful of callbacks and I'm through to the coding round at Meta. Y'all are the best. thanks for the encouragement and strategy here. Leetcode time.
Also, to anyone reading this, the LinkedIn EasyApply button is wildly overpowered lmao.
Would you be open to sharing how many callbacks you got from it? I've never heard back from easy apply
Sai - I easy applied to 120 jobs and have 10 followups, 10 rejections, and haven't heard back from the rest.
Ooo that's awsome. I've been sleeping on it. Thanks so much for sharing!!!
Sorry to hear about the layoff and the added anxiety that comes with dwindling savings and supporting your family.
My first research paper ever was published at NeurIPS last year.
Congrats! This seems like such a huge advantage over almost all other AI candidates -- you have a paper published in the preeminent AI conference in the field!
I encourage you to focus on AI roles for 3 reasons:
One nuance: with Big Tech (Meta/Google), I'd potentially be flexible about what role you take if they express a strong desire for that. Since it sounds like these companies are at the top of your list, optimize for getting in, and you can likely switch into an AI team eventually (or at least an AI-adjacent team).
Regardless of whether you get a Big Tech offer or not, I'm very confident you'd get some offer. Lots of startups would love to hire you, so don't split your attention by applying to eng manager roles.
So to answer your question directly, I'd focus on #1 (Go all out for AI Engineer Roles (passion forward)) and consider #2 (non-AI roles) for the Big Tech roles!
thanks! feeling better already.
Sorry to hear about the layoff 😢. I just want to start off by saying that careers are long (I'm assuming you're not 60 years or something), so even if you don't end up in the role you want now, you can always pivot into it later. Given that you're a Staff Engineer/Engineering manager, I imagine a lot of your skills are transferable too, making pivoting easier (and given your level, people will have more trust in your ability to learn fast).
I agree with trying out all 3 paths. I would do the following:
Digital ink is cheap, so if you really grind at it, you can easily do this in 1 week, max 2.
After you do this, see where you're finding traction and double down. This is why it's important to construct funnels and track your conversion rates as I talk about in my job searching course here: https://www.jointaro.com/course/ace-your-tech-interview-and-get-a-job-as-a-software-engineer/organizing-your-job-search/
So let's say your conversion rate for #1 (AI/ML) is 2 interviews (4%), for #2 (web) it's 10 interviews (20%), and for #3 (EM) it's 3 interviews (6%). In this case, it's obvious that to me given your constraint of needing something in 3 months to just drop #1 and #3 and go all-in on #2 (web dev). Studying for 3 different types of roles will almost certainly drive you crazy and lead to mediocre performances for each one.
thank you, I love the data-driven approach!
I know this wasn’t the point of the question, but even if you are 60 careers can still be long. My dad retired last year from being a senior IC at Intel at 86. So I think the goal is to really find something you love doing and keep doing it for fun! He actually still talks to his team every week about their projects… but that’s so much better then puttering around the house with no purpose!