Meta E6 Menlo Park Onsite Experience (Rejected)
Summary
I interviewed for a Specialist AI role (research engineer) at Meta's Menlo Park office and was ultimately rejected after an onsite experience that included coding, AI-specific coding, AI design, and behavioral rounds.
Full Experience
Giving back to the community.
Specialist AI role (research engineer in a specific field). The interview process seems to be different from general software or machine learning roles.
First, recruiter did a team "pre-match" to identify 3 or so teams that match my background.
Then 2 phone calls with potential team members with informal chats about my resume & background. There were no "interview" questions at all (no coding, ML knowledge, system design, behavioral etc.)
Onsite:
- Coding: Two LC mediums, both in Meta top 100 (around 70 or 80).
- AI Coding: One LC easy, and one about PyTorch (for a single layer network, give high level description of forward() and backward())
- AI Design 1, 2: Both were questions about my specific field. Describe a system on Excalidraw and discuss various aspects, citing different approaches used in the field. Totally different from general Meta ML design interviews on youtube/books.
- Behavioral: Usual questions you can find on the web. Describe conflict with coworker & how you handled it & what you would do differently if you could go back.
After interview, hiring managers of the pre-matched teams looked at the interview feedback to decide whether to send my packet to candidate review or not. Apparently, hiring managers decided to reject. I had to wait almost 2 weeks to hear results because recruiter also tried asking other hiring managers.
Interview Questions (2)
For a single layer network, give a high level description of forward() and backward() methods in PyTorch.
Describe conflict with coworker & how you handled it & what you would do differently if you could go back.