Which data role owns a task: data engineer vs database administrator vs data analyst
Verdict: Match the task to what it touches. Moving, cleaning, or transforming data between systems is the data engineer. Securing, backing up, restoring, or tuning a database is the database administrator. Turning processed data into insights is the data analyst.
| Criterion | Data engineer | Database administrator | Data analyst |
|---|---|---|---|
| Owns | Ingestion, cleansing, and transformation pipelines between systems | Database security, availability, performance, backup, and recovery | Turning processed data into insights, trends, and decisions |
| Typical tools/tasks | Designs and monitors pipelines that load a warehouse; builds cleansing routines | Grants or denies user access; restores from backup; tunes a slow database | Explores raw data; builds analytical models, charts, and reports |
| Not responsible for | Granting access or restoring backups | Building transformation pipelines or visualisations | Managing pipelines or database security |
Rules
- If the task is granting/denying access, backup and recovery, or keeping a database available and performant, the database administrator owns it.
- If the task is designing ingestion, cleansing, or transformation pipelines that move data between systems, the data engineer owns it.
- If the task is exploring data to build charts, visualisations, or reports for decisions, the data analyst owns it.
- Tuning a production database so it stays fast and reachable under load is a database administrator responsibility, not a data engineer one.
- Managing and monitoring data pipelines so loads perform as expected belongs to the data engineer, not the database administrator.
- Designing analytical models and processing raw data into business insights is the data analyst's job, distinct from the pipelines that deliver the data.
Traps
- A data engineer preparing clean data is not the same as an analyst exploring it: the engineer's job ends at delivering clean data through pipelines, not producing trends or visual reports.
- Being close to the stored data does not make the database administrator the pipeline builder: the DBA secures and maintains the database, not ingestion and transformation pipelines across systems.
- Consuming warehouse data does not make the data analyst the pipeline owner: analysts explore and visualise data rather than building the pipelines that load it.
- Requesting or using data access is not administering it: only the database administrator grants or denies user privileges over the database.