Data Quality Intern (d/f/m)

Role Overview This internship (three to six months) involves discovering data quality issues in our part and component data, reframing them into customer‑relevant problems, measuring impacts, and delivering evidence‑backed findings to the larger product team. The role emphasizes data, analysis, and scripts to drive fixes from discovery through production. Responsibilities and Objectives Turn ambiguous data‑quality questions into customer‑relevant findings by reframing part‑level observations into business or customer impact, defining a sensible metric or proxy, and producing a measured, caveated answer to the question set by the product manager. Independently size problems and test hypotheses against data by writing read‑only queries over ClickHouse and PostgreSQL, producing defensible numbers while recognizing double‑counting or misleading results. Make the effect of fixes and experiments visible by extending dashboards and building ad‑hoc visualizations that show trends, baselines, and whether an intervention moved coverage or correctness. Run small experiments to gather evidence by writing scripts (with AI assistance) against external sources such as SiliconExpert and DigiKey to check gaps, calibrate findings, or perform spot checks. Verify product assumptions by navigating the epibator (Rust/TS) codebase with AI tooling to confirm how data is resolved and, if needed, add light instrumentation without owning large refactors. Apply scoped fixes safely by writing AI‑assisted scripts that correct production customer data at scale, using dry‑runs, sample validation, reversible checkpoints, and guardrails to avoid regressions or data corruption. Leave behind durable, trustworthy knowledge by following the mission loop—brief, investigate, report, distill—citing evidence, dating facts, and writing findings for other teams to act on without re‑deriving them. Act as your own harshest critic, reconciling and sanity‑checking results, separating what is proven from what is still a hypothesis, and flagging loudly when a finding overturns a prior assumption. Qualifications and Skills Pull trustworthy numbers out of messy data with SQL. Own the question, not just the query, and reframe it toward what truly matters. Sanity‑check work and truthfully acknowledge uncertainty. Communicate findings clearly enough for a non‑analyst to act on. Use AI with real verification, and write small Python scripts to fix data safely. Show high agency, learn quickly, and avoid dropping threads. Bonus: read code with AI help and be curious about the electronics domain. Deep technical and domain depth is coachable; calibration and judgment are not. Technical Stack ClickHouse as data warehouse and PostgreSQL for production data. Python for scripting, data transformation, and safe bulk fixes. Rust and TypeScript in the main product codebase; read code with AI help to verify behavior. External data sources such as SiliconExpert and DigiKey accessed via APIs. AI tooling (e.g., Claude Code) for code navigation, querying, and scripting. #J-18808-Ljbffr Luminovo