Taro Logo

Pondering a transition to AI/ML but have a fear of failure - What to do?

Profile picture
Anonymous User at Taro Community6 months ago

Hey guys, please guide me for the next steps in this journey.

I am currently L5 at social media company (not Meta) and have done Android dev for most of my career. I was a manager for a brief amount of time as well but layoffs took that away. I am on a H1B visa and family is dependent on me (single earner).

From last 4 years or so I haven't done much dev work and have been working with developers/customers/partners to help them in integrating our solution to their Android apps. I can crack LC any day though but might not be able to pass Android round easily. I don't enjoy Android dev too much now but it's a different ask if it becomes a necessity to survive at this level. I used to be really good at it and if I put 3-4 months to it I can pick it up again. I am interested in AI/ML given the popularity, high TC, etc. but also interested because I genuinely believe that it can transform people's lives + it will broader my skillset beyond tech companies and gives me a long run. But I have almost 0 experience in it.

I fear another layoff at my company and it may impact me in next 6 months, or it might not happen either if market or company situation improves.

Question now is: Should I get back to my Android roots OR try to dip my toe into AI/ML? If latter then does Certification course from Standford/Berkley,etc. helps?



  • 3
    Profile picture
    Tech Lead/Manager at Meta, Pinterest, Kosei
    6 months ago

    Choose the shortest path back into an active coding role.

    You mentioned that you haven't done much dev work in the past 4 years or so, so I'd first focus on remedying that. If you're a Partner Engineer or Solutions Engineer, the barrier to be a Software Engineer is a bit higher -- it may require re-interviewing even within the same company.

    It sounds like you have great Android experience in the past, so I'd use that to transition in to a mobile engineer role. Once you're in that position, it'll be easier to transition to AI/ML, which is your eventual goal.

    I don't think a certification helps in these cases. They are expensive and take time. You work in Big Tech so you presumably know lots of people; can you use them for referrals?

  • 2
    Profile picture
    Tech Lead @ Robinhood, Meta, Course Hero
    6 months ago

    Since you're a single earner on a visa, I highly recommend taking the most consistent and stable path: Sticking with Android and becoming extremely good at it again. Since you're an L5 engineer, I'm sure there's a lot of fundamentals you have built up that will come flowing back once you break out the Android Studio again.

    From there, you can run the playbook Rahul described:

    1. Get back into coding - Establish yourself as a strong L5/L6 coder again with Android
      1. After you get back into it, see if you enjoy it or not. If not, you can start thinking about a pivot. If it turns out that you actually love it again, you might want to just stay in Android! We made a video about this process here: How To Discover Your Work Passions And Hatreds
    2. Transition internally to MLE - Since you work at a Big Tech company, I'm sure there's a lot of ML roles. Internal transfers are also way better due to the trust you've built up within the company, and you'll have a leg up from understanding the company culture. It is way harder to do a pivot via interviewing - It's too much risk for the hiring company.
    3. Do whatever you want as an MLE - Stay in your company, move to another company, or become an MLE manager.

    Another thing I want to say is: It's way more financially lucrative being in the Top 5% of something than the Top 25%. Since you already have an Android background, it's much easier for you to be in that Top 5% and command top TC as an expert specialist in the craft. With MLE, it's way harder as you're starting from scratch and you now need to compete with the millions of other smart engineers who also want to switch into MLE off of the AI hype.

    Here are some other resources on transitioning to ML: