About
mlmentorship
Senior ML & AI interview prep, by Hamidreza Saghir. Free.
Author
Hamidreza Saghir, Principal Applied Scientist at Microsoft. Earlier roles include Senior MLE Lead at X (Twitter), Applied Scientist at Amazon, and ML Researcher at Borealis AI (RBC). PhD from the University of Toronto, M.Sc. from Sharif University of Technology. Publications at ACL, InterSpeech, The Web Conference, IEEE TASLP, and Physical Review E.
I keep my opinion writing, consulting, and Looplet (an iterator-first agent framework) under my own name at hsaghir.com. mlmentorship is the focused interview-prep companion.
Why this site exists
Friends and ex-colleagues kept asking me the same questions about senior ML jobs and interviews: how to prep for the loop, how to think about L5 vs L6 calibration, what evals to build for an LLM product, how to scope an ambiguous applied-ML problem. After answering them in DMs and 1:1s a few too many times, I started writing the answers down. mlmentorship is the version I felt comfortable putting on the public web.
Public ML interview prep is calibrated for new grads and bootcamp-to-FAANG transitions. Free content for that audience is abundant.
Less common is content for mid-career ML practitioners targeting L5+ roles. The questions are different, the bar is different, and the signal interviewers read is different.
This site is opinionated, specific notes on what senior Applied Scientist, MLE, and Research Engineer interviews actually test for, and on the topics that come up most often.
What's here
- Start here: curated reading paths by what you're trying to do.
- Questions: the senior ML canon, with what L4 / L5 / L6 answers sound like.
- Guides: long-form pieces on senior ML interviews, system design, and patterns from teams that ship.
- Concepts: 99 concise notes on attention, normalization, LLM internals, optimization.
What's not here
- Generic ML basics. The internet has this covered.
- New-grad / bootcamp / "break into ML" content.
- Promises about landing a specific job in a specific timeframe.
Reach me
- Personal site: hsaghir.com (essays, consulting, Looplet)
- Twitter / X: @hrsaghir
- LinkedIn: linkedin.com/in/hamidrezasaghir
- GitHub: github.com/hsaghir
- Email: ml.mentorship [at] gmail [dot] com