Recruiter screen, coding, ML breadth, system design, depth, behavioral, team match — not a generic four-stage sales pipeline.
Every scheduled round names the material it covers and links to the modules that prepare you for it.
Write down what was asked while it is fresh. Four loops later the recurring weakness is on the page instead of in your memory.
Base, equity, refresh, level and start date compared honestly, with the deadlines that force the decision.
Every company, every round, every deadline on one board. Overdue follow-ups surface on their own instead of being remembered at midnight, and a stalled process looks like a stalled process rather than like silence.
A system design round on Thursday puts the ML System Design Foundations module in front of you on Monday. The tracker and the curriculum are the same product, so preparing is never a separate act of organization.
Questions asked, what you answered, what you wish you had said. It is the raw material for the next loop, and the only reliable way to learn that your weak round is system design and not coding.
A filter, looking for two things: that your experience matches the posting closely enough to justify the team’s time, and that you can describe it without rambling. Have a ninety-second version ready.
Either standard algorithms, or the ML variant where you implement k-means, attention or beam search from memory. Most candidates prepare for the first kind only.
Rapid questions across the whole field. What is graded is calibration: two sentences when that is what was asked, three levels deeper when the follow-up wants it.
The round that decides your level. Running out of time before you reach the tradeoffs is the most common way it is lost, which is why the tracker budgets the sections for you.
Graded, and vague answers cost offers. Specific stories with real numbers, each one able to survive three follow-ups.
