Data Engineer
Build traceable vehicle-data pipelines that preserve identifiers, source dates, coverage and incomplete records.
Car Check helps people understand a used car’s history. We bring vehicle records together with clear explanations, keeping the source, timing and limits of the information close to each finding.
We care about the whole experience, from the reliability of a lookup to the words that help someone decide what to ask next.
The opportunity
We are looking for a Data Engineer to make the information behind Car Check consistent, traceable and useful. Vehicle records arrive with different definitions, dates and gaps. Turning them into a report requires careful decisions about what to preserve, what to normalise and what we still do not know.
You will work with backend engineering and vehicle research on the path from a source response to the fields a buyer sees. The goal is not simply to move data between systems. It is to make it possible to explain where a value came from, why it changed and whether it is appropriate to compare it with another record.
You will help establish dependable data models and quality checks while keeping the work practical. We want a system that makes discrepancies visible early and gives the team a clear way to investigate them, rather than relying on a growing collection of silent exceptions.
What you’ll work on
- Build repeatable ingestion workflows. Validate incoming records, preserve the relevant source context and make it possible to rerun a failed step without introducing duplicates or losing history.
- Define useful data models. Work through identifiers, dates, units and relationships with engineering and research so that a field has the same meaning throughout the product.
- Make quality measurable. Create checks for missing fields, unexpected values, freshness and changes in source formats, with clear routes for investigating the findings.
- Preserve provenance. Keep enough information to trace a report value back to the source and transformation that produced it, including when the information was obtained.
- Handle corrections and reconciliation. Design safe ways to compare conflicting records, revisit transformations and update derived results without obscuring what changed.
- Support useful analysis. Provide documented datasets and queries that help the team answer product and coverage questions without confusing collection gaps with customer behaviour.
What you’ll bring
- You have built data pipelines or transformation workflows that other people depended on, including handling failures, backfills or changes in upstream formats.
- You are comfortable with SQL and a general purpose language such as Python or TypeScript, and can explain the tradeoffs behind a data model you designed.
- You investigate discrepancies methodically, checking definitions, timing and lineage before treating an unusual number as either an error or an insight.
- You understand the value of validation, reproducible transformations and documented assumptions, especially when data arrives incomplete or changes over time.
- You can communicate limitations clearly to product and research colleagues and help them decide what the available information does and does not support.
Useful, but not required
- Experience with entity matching, event histories or datasets assembled from several independent providers.
- Work on data quality monitoring, lineage or tooling that helped other teams investigate unfamiliar records.
- Familiarity with regulated information, vehicle datasets or products where the date and source of a finding matter.
Your first few months
- Trace a representative set of report fields from source to display and document the assumptions or transformations along the way.
- Improve a data workflow with repeatable processing and checks that expose a meaningful class of quality issues.
- Give the team a documented way to investigate a discrepancy and understand the effect of a source or transformation change.
How we work
Start with the buyer, follow the evidence and make the complex clear. We value thoughtful questions, shared learning and care in the details of what we build.
We work in focused pieces, make important decisions easy to follow and bring questions into the open while there is still time to act on them. You will have room to shape the work, with the responsibility to explain your choices, ask for help and follow through on the result.
Apply for this role
Share an example of a dataset or pipeline you improved. Explain a quality issue, how you traced it and how your changes helped someone rely on the resulting information.