This role sits within our EMEA, Data and Technology team.
In this role, you will own and extend our config-driven data platform (DMI), which standardises ingestion, transformation, and delivery of paid media data across 26+ ad platforms for multiple global clients. You will work closely with our team to build and maintain ELT pipelines from Cloud Function ingestion into BigQuery through to dbt-powered transformation ensuring the highest standard in data integrity and scalability.
This is an exciting role with excellent career opportunities within a high-profile team and scope to strategically shape the agency. We are looking for someone who can hit the ground running, contribute to a mature mono-repo data platform, and help drive best practices across the engineering team. Experience with digital media data is highly beneficia
Responsibilities
- Own and extend the end-to-end data pipeline from Cloud Function ingestion through dbt transformation (staging → intermediate → marts) to analysis-ready tables in BigQuery
- Develop and maintain dbt macros, Jinja templates, and platform YAML definitions that auto-generate models across 26+ ad platforms
- Manage and improve GCP infrastructure (BigQuery, Cloud Run, Cloud Functions, Cloud Scheduler, Pub/Sub) provisioned via Terraform.
- Build and maintain the Python CLI tooling that orchestrates client onboarding, config compilation, and pipeline execution
- Mentor the team of data engineers, driving best practices in DataOps, code review, testing, and documentation.
- Proactively review existing processes to identify opportunities to automate manual work, optimise data delivery, and re-design infrastructure for greater scalability
- Collaborate with analysts, data scientists, and BI teams (PowerBI, Looker Studio, Tableau, etc.) to maximise the value delivered from data mode IS.
- Contribute to CI/CD pipelines (Cloud Build), testing (pytest, dbt tests), and documentation (MkDocs, etc).
About You
Required:
- Strong experience with dbt – macros, Jinja templating, incremental models, seeds, testing, and packages.
- Proficient in Python 3.11+ building CLI tools, data processing, and automation.
- Proficient in SQL, ideally BigQuery dialect.
- Experience with Google Cloud Platform especially BigQuery, Cloud Run, Cloud Functions, Cloud Storage, Pub/Sub, and Cloud Scheduler
- Experience with Infrastructure as Code (Terraform) for provisioning and managing cloud resources.
- Solid understanding of data modelling techniques (star schema, dim/fact architecture, slowly changing dimensions)
- Comfortable with Git (GitHub, branching strategies, pull requests) and CI/CD (Cloud Build or similar)
- Ability to translate business needs into technical specifications.
Highly Desirable:
- :Experience with Docker and containerised workloads (Cloud Run Jobs.)
- Familiarity with CLI frameworks (Click) and config-driven architectures (Pydantic, YAML-based configuration)
- Knowledge of the digital media / paid media industry — we process data from 26+ ad platforms (Google Ads, Meta, DV360, TikTok, etc)
- Exposure to multi-cloud integrations (Azure Blob, AWS S3, SFTP)
- Mono-repo experience — managing multi-client configurations in a single codebase.
Nice to Have
Experience with Databricks (and dbt-databricks.)
- Familiarity with modern Python dev tooling — Poetry, ruff, mypy, pre-commit.
- Experience with docs-as-code (MkDocs or similar).
Qualities:
- Ownership – an ability to manage multiple workstreams across clients with accuracy, and see things through from design to deployment.
- Curiosity – a natural inclination to explore new tools, dig into unfamiliar systems, and understand how things work end-to-end.
- Resourcefulness – an ability to unblock yourself, whether that means reading source code, querying logs, or finding creative workarounds when data or documentation is limited.
- Problem-solving – an ability to think through complex data issues methodically and design clean, maintainable solutions.
- Collaboration – a desire to work openly, share knowledge, and build a team culture where code reviews and pair programming are valued.
#J-18808-Ljbffr…