Live Coding: SQL & Python

Can you translate a business question into correct, performant SQL (joins, window functions, CTEs, dedup) and write clean Python for data manipulation — while narrating your reasoning. Weak SQL is the single most-cited rejection reason across companies.

• Drill window functions (RANK/DENSE_RANK/ROW_NUMBER/LEAD/LAG), top-N-per-group, running totals, gaps-and-islands.
• Practice dedup and "latest row per key" and explicit > vs >= boundaries.
• Write Python for parsing/transforming messy data; know pandas groupby and exception handling.
• Talk out loud: state assumptions, handle NULLs and edge cases before optimizing.

• Forgetting NULL handling and duplicate keys.
• Reaching for self-joins where a window function is cleaner.
• Silent coding — interviewers want the reasoning, not just the answer.