Position: Senior Data Delivery Engineer
Mode: Fully remote
About the role:
You will own and operate our dbt transformation layer that converts raw transaction data into clean, consistent, and trustworthy datasets for internal products and client delivery. This role is hands-on: you will write and maintain dbt models and rules, implement automated validation, monitor data quality, and troubleshoot issues quickly as upstream data and business logic change.
Responsibilities:
- dbt-first transformation ownership (models, tests, documentation)
- Receipt data cleaning rules (normalisation, standardisation, edge cases)
- Data integrity (quality gates, monitoring, incident response)
Problem space
Receipt data is inherently messy. You will routinely work through:
- Retailer and format variability, missing/ambiguous fields, and inconsistent line items
- Edge cases and exceptions where rules must be explicit and versioned
- Schema drift and upstream changes that can silently break assumptions
- Data quality incidents (freshness/completeness/correctness) requiring fast triage and durable fixes
Skills & experience
Must-have
- Strong SQL and practical data modelling skills (staging marts / delivery outputs).
- Production experience with dbt (models, tests, docs; comfortable with refactors).
- Solid understanding of data warehousing and ELT/ETL concepts.
- Experience working with a cloud data platform (AWS, GCP, and/or Azure).
- Strong problem-solving and debugging skills; high attention to detail.
- Clear communication and ability to collaborate across product/ops/engineering.
Nice-to-have
- Python for data tooling/automation and future pipeline work.
- Experience with messy transactional datasets (e.g., receipts), schema drift, or semi-structured sources.
- Data observability/quality tooling experience (custom monitoring, Great Expectations, or similar).
- Orchestration and CI/CD exposure (Airflow/Dagster/Prefect; PR-based release workflows).
Responsibilities:
1) Own dbt models and cleaning rules
- Build, maintain, and improve dbt models that clean and standardise receipt data.
- Translate business rules into durable transformation logic (including handling edge cases and retailer variability).
- Keep the project maintainable through refactoring, consistent patterns, and clear documentation.
2) Data quality, validation, and integrity
- Implement and maintain automated tests (schema, uniqueness, not-null, relationships, accepted values) and custom tests where required.
- Define quality expectations for key outputs and ensure failures are caught early.
- Investigate anomalies, identify root cause (source vs model vs rule), and implement durable fixes.
3) Monitoring and operational reliability
- Monitor scheduled runs and downstream outputs (freshness, completeness, key metric sanity checks).
- Improve observability and incident response: alerts, dashboards where appropriate, and runbooks for common failure modes.
- Reduce operational toil by addressing recurring issues systematically.
4) Collaboration and delivery alignment
- Partner with stakeholders to clarify requirements and ensure datasets meet delivery needs.
- Communicate trade-offs clearly (accuracy vs coverage vs latency vs complexity).
- Review contributions to the dbt project and help raise engineering quality across the team.
5) Performance and cost discipline
- Optimise model performance and warehouse usage (incremental strategies where appropriate, efficient joins, reduced scans).
- Balance correctness, coverage, and cost constraints.
6) Data ethics and privacy
- Follow data privacy and responsible handling requirements.
- Identify potential compliance risks and escalate or propose mitigations.
Tools (working environment)
- Snowflake, PostgreSQL
- Git-based workflow (PRs, reviews, CI)
- Cloud platform exposure (AWS preferred)
