ML Career Questions: Pathway to Senior ML/DS?
If you're early career in ML/DS/AI, what should you know to plan your future? Sharing my answer.
Reader question:
I think it would be very useful to learn about the career path progression in ML/DS. The criteria/milestones that usually make the path progression. Do these thing usually happen in the same company or happen in transitions? What are the pros and cons. Things that are not easy to find written somewhere and that are not spoken about often.
Susan:
Great question, I think this is often hidden information that I only learned about from people more senior than me sharing them candidly when I was early career.
Company size has a big impact on leveling.
Larger companies tend to have more established levels, with rubrics. An example that's shared externally is this Etsy engineering ladder (lots of other companies have also shared)
https://etsy.github.io/Etsy-Engineering-Career-Ladder/
but internally this can be a very detailed grid. Most large companies I know have this, and people interested in promotions go and make sure they are doing some things on the next level of the grid (while covering their current role)Startups without things like this rely more on relationships and ownership of projects (everyone, including managers, are trying to figure things out, including leveling)
There are less openings the higher the level. If someone is already doing that technical lead role, there might not be a reason for someone else to be promoted into it for the time being if the company/team doesn't need 2.
Lots of luck is involved. People could take 2-3 cycles or more to get promoted, if the way the promotions are decided are something like this: your team has 10 people, and there are 2 sister teams, total 30 people. In these 30 people, budget or procedure is approved for 1 promotion. The 3 managers of those 3 teams make the case for their candidate. In the end, one of those candidates get the promotion that cycle; it might be you, it might not. (Usually bigger companies)
Salary vs. title: I know someone that went from principal at a large company to senior at a growing startup who paid more. Personal choice
This was a good read from Charity Majors, who writes great posts on management and IC (individual contributor) roles and growth:
https://charity.wtf/2020/09/14/useful-things-to-know-about-engineering-levels/
See the original thread here: (Linkedin.com)
Even if you’re early career, just knowing about these systems can shave 3+ years off your career growth. Make sure you’re learning fast and often especially early career, and don’t be afraid to take risks. Risks I took early career include not specializing too soon; I was able to ship recommender systems but also reinforcement learning. Then I moved to time series. I’m at a point where I’m thinking of specializing more, but the fact is those risks I took to learn something from scratch again helped me grow so much.
Hope this was helpful! Feel free to leave comments and questions.