Wrapped the VMware data scientist loop in January 2026 for a role on their customer success analytics team. This was my first enterprise-infra DS interview and it had a different texture than what I'd seen at consumer or fintech companies. Sharing the breakdown.
SQL round: Definitely present. One dedicated 45-minute SQL round. HackerRank or a shared doc, I did mine in a doc. Questions were intermediate to advanced. Window functions came up immediately: running totals, ranking within groups, lag/lead for month-over-month comparisons. One multi-table join with aggregation. Not just "select from where group by" stuff.
They also asked me to interpret a slow-running query and identify the bottleneck. That was a bit unexpected. Have at least a passing familiarity with query execution plans and indexes.
Statistics / methods round: About 30 minutes of this within a longer round. Topics: A/B test design, how to determine sample size, how to handle imbalanced groups, how to present results to a non-technical VP. One question about cohort analysis for customer retention. They did not ask me to code a model live. It was more conceptual.
Case / product sense: "A key customer segment's support ticket volume has risen 40% over the last two quarters. Walk me through how you'd analyze this." This felt like a mix of DS and PM. They wanted data sources, hypotheses, an analysis plan, and a narrative for how you'd present findings to leadership.
What I didn't see: Machine learning depth (no model architecture questions, no feature engineering deep dives). This role was analytics-heavy, not ML engineering-heavy. Know your role type before you prep.
Comp data point: offer was for remote, mid-senior DS. Base around $155k, bonus target 10%, RSU grant in Broadcom stock over 4 years. Total first-year closer to $185k with vesting.