[Microsoft] Applied Scientist loop: more classic than the labs, but the ML depth is real

Microsoft MLE / Applied Scientist felt more like a structured big-tech loop than OpenAI/Anthropic: recruiter, an online assessment, then onsite with DSA coding, an ML breadth/depth round, and ML system design.

Less "vibes," more checklist. If you're coming from a startup, the bar is predictable, but they do probe fundamentals seriously (bias/variance, metric choice, how you'd actually deploy and monitor). Know your basics cold; the curveballs are fewer but the fundamentals bar is real.