If I don't like the domain at my new job, how difficult would it be to switch to another one (particularly, machine learning)?

Overall it usually depends on a specific company. E.g. in most of FAANG companies it's pretty easy to switch between different teams and even departments. However in other companies a lot depends on the leadership. I would recommend to start chatting with people you trust within your current company. And if you were able to build a good relationship with your direct manager, that could be the best opportunity to discuss where you want to grow into your career.

I'll start off by saying that it's way easier to switch stacks within a company than by doing it via a job switch. This is because the company you're switching into will be very concerned about your raw productivity since you'll be new (and they probably gave you nice pay raise to join them). Productivity will obviously go down if you're doing X at your current company and then joining the new company to do Y.

On a side note, I was able to pivot stacks via a job switch as I switched from a back-end engineer role at PayPal to an Android lead role at Course Hero. Here were the 2 core factors:

1. My side projects - At the time of switching, I had published multiple Android apps for fun across ~2 years that had ~50,000 users combined. I don't think this angle is as applicable to ML.
2. I got offers from startups - I feel like startups are more amenable to hiring "scrappier" talent (like me in 2015). ML is quite hot nowadays, so I imagine this angle does apply.

That being said, here are 4 factors that go into ease of switching stacks:

1. Your relationship with your manager
2. Culture around job switching at your company
3. Availability of opportunities within that stack in your company
4. Your seniority

I don't know much about Rubrik, so I can't comment much on #2 and #3, and for #1, if you want advice there, ask questions into Taro!

So let's dive deeper into #4.

I strongly believe that switching stacks is much better the more senior you are with the turning point reaching a senior engineer level at a strong company (which I believe Rubrik is). The underlying reason behind this is that your technical fundamentals (i.e. engineering instincts that are universally transferrable) get better when you go deep into 1 thing, which is generally what senior engineers do. Zooming out from that, senior engineers have stronger non-technical fundamentals as well like getting better at 1 on 1 meetings, building relationships, having the confidence necessary to ask for what's important faster, and processing lots of information efficiently. This all leads to increased learning speed, making stack transitions far smoother.

All that being said, I recommend "failing fast" when it comes to figuring out if you inherently hate a stack of work and taking 3-6 months as opposed to a year. If you like your manager, keep them closely in the loop as you're figuring this out!

Related resources:

• This Is How Software Engineers Should Initially Learn
• "How can I write great code faster?"