Background:
Unlike most tech companies that recruit candidates for specific teams, new employees at Meta need to join a bootcamp program. While finishing up many online trainings, random technical coding tasks, and setting up firmwide logistics, new employees are required to look out for new job openings from the company’s Jobs Tool similar to Linkedin’s jobs panel . Ideally, a new employee should find a team to join within 6 to 8 weeks to graduate from bootcamp. I have been told that the bootcamp also serves as an unofficial probationary period, some unknown actions may result a HR talk.
Challenges:
- There were over 1,000+ new tech joiners in July globally. There were 13 Machine Learning Engineers, 21 Research Scientists, and total 211 new engineering joiners in the Greater Seattle area. However, there were only less than 20 openings for Machine Learning roles.
- There were too many new hires in 2021. Many teams are full of newbies who have less than 2 years of experience.
- During COVID, 2020 and 2021 employees rating were massively inflated. There were too many E5 took shortcuts to be promoted to M0 then became M1. It will be risky to join a team led by such hiring manger.
My Approach:
My principles:
- Follow your focus areas and/or interests, not what’s trendy in the company/industry, i.e., VR, Metaverse, Deep Learning, etc.
- Follow the money, not the company’s risky goals.
- How does the company actually make money? Selling ads. Then join the team that help selling more ads and making more money out of ads.
- Do look around other people, may consider follow superstars, do avoid low performers.
- This may be the last chance to ask uncomfortable questions:
- During the manager-talk phase, don’t be afriad of asking bitter questions or making them feel pressured.
- Ask the manager what are their strengths and weaknesses, just like how they asked you during your interviews. Verify their answers with their direct reports.
- Scarcity has its leverage. Join the team that does not have too many similar people like you.
The Jobs Tool will only be open to new employees after their first 3 weeks. Within the first 3 weeks:
- Talk to trustworthy friends who have good growth track records and also know your strength or interests.
- Use internal tools to find all of your acquaintances, alumni, ex-coworkers.
- If they showed up in your past life, you may have similar qualities.
- Sort these people into 3 categories: superstars, peers, losers/coasters.
- Consider join the team that superstars work at.
- Consider join the org or business functions that your peers work at.
- Avoid the org and business functions that losers/coasters work at.
After the Jobs Tool opens up, use your focus areas and interests as top priority to find openings. As of July 2022, Meta’s openings Machine Learning employees are mostly related to Recommendation and Ranking algorithms. Here are my thoughts:
- Recommendation and Ranking has been a 20+ year popular topic. Too many people know about it. This will be a competitive and low margin area.
- The company uses convoluted deep learning models, which are tuned in Meta AI originally “FAIR” . Most of Recommendation and Ranking product models are developed on top of those root models.
- The amount of profit that Recommendation and Ranking can bring to the company needs more effort to justify than build.
Pros and Cons of the Team Selection process
- You have time to try out different teams, but the team you like can also be picked by other bootcampers.
- When receiving the offer, you will not know if there exists a team that you like.
- Managers may sell you a good impression, but it’s difficult to verify. A manager told me that his Pulse Score a score given by mean ICs, max 100. from last quarter was 98.. I was sold until someone told me that this manager joined the team recently and the score actually was for the ex-manager at this position.
- The 8+ weeks bootcamp time is relatively chill, but can be a wasteful of time.
But, you can’t predict everything, especially the people
- It’s impossible to choose who you are going to work with. But you can assume most of engineers are introvert.
- Meta’s frontline engineering manager has also became an unique and somewhat infamous role in the tech industry. That is, they do not need to contribute on the technical side of the team. As a E7 friend who has been in the company for 8+ years said, M1 seems to be an individual contributor who’s sole job is to take notes for others' PSC.
- Based on my limited exposure (talking to ~20 employees) to Meta’s culture, it tends to undermine employees relationship and underscore project impact. Yet, for senior and above employees, the key of sucess is from the collaboration with people.
Now what?
Team selection comes with lots of effort. It all boils down to the principles that I made before the selection process, a set of goals and thing that I want to achieve and accomplish within the next 12 months. Therefore, when I reach the goals, or when things derail the plans, that will be the time to bounce.