JPMorgan Data Analyst Case Interview: Transaction Risk & Client Insights Dashboard

This case mirrors JPMorgan’s client-first, controls-minded culture and the scale of its Corporate & Investment Bank and Asset & Wealth Management data environments. Candidates receive a realistic, messy dataset (e.g., trades/transactions, client reference data, KYC/AML flags, market data snapshots) and a brief from a non-technical stakeholder asking for a daily dashboard that surfaces client flows, revenue attribution, risk hot spots, and potential suspicious activity. The session assesses structured problem solving, data quality rigor, and stakeholder communication under compliance and audit expectations. Focus areas: (1) Clarify business metrics and controls—define consistent KPIs (net inflows/outflows, exposure by sector/region, realized vs. unrealized P&L, client profitability, alert rates), document assumptions, and state data lineage and retention considerations. (2) Data profiling and governance—identify nulls, late-arriving trades, timezone and currency normalization, deduping logic, reference-data joins, and propose validation checks (reconciliation to prior day totals, outlier thresholds, sampling for audit). (3) SQL-first analysis—write or outline queries using window functions, conditional aggregation, and joins across fact and dimension tables to compute KPIs and triage data quality issues; discuss performance trade-offs and indexing for large tables. (4) Python/pandas reasoning—sketch how you would detect anomalies (e.g., sudden spikes in client flows or unusual instrument activity), handle edge cases, and produce a tidy dataset for visualization; discuss reproducibility and code review expectations. (5) Visualization and narrative—design a concise dashboard (e.g., client flow waterfall, heatmap of alerts by sector/region, top client contribution, exception queue) tailored to a risk/controls audience; justify chart choices and alert thresholds. (6) Risk and regulatory lens—explain how the pipeline would meet JPMorgan’s controls (auditability, access management, PII handling, separation of duties), and how you would document metrics and data lineage for model/risk review. (7) Stakeholder communication—deliver a brief readout translating technical findings into business actions (which clients/segments to investigate, which data issues to remediate today vs. later) and propose an incremental rollout plan (MVP, data quality SLAs, monitoring). Evaluation emphasizes clarity of assumptions, accuracy and consistency of metrics, depth of controls thinking, and the ability to balance speed with rigor in a highly regulated environment.