Snowflake Data Engineer

Job

Welcome to Group Technology, where we pride ourselves on engineering solutions and direct Nordea’s transformation by providing a holistic technological view and structured understanding of the bank, and its surrounding environment to enable the Customer Vision and the Business Strategy.

Nordea is a place where traditions meet tomorrow. We're not just a bank, we're a tech employer on a mission to evolve finance securely and responsibly. Together, we impact millions of people’s daily lives by ensuring they can access our solutions anytime, anywhere, while safeguarding their personal data and wealth. Join us in making an impact on the banking industry.

You’ll join a Large Data Foundation team in a banking environment responsible for building and maintaining transformation layers on Snowflake — turning ingested raw data into clean, governed, analytics‑ready datasets consumed for IRB risk Model development and Model use for risk parameters.

We work cross‑functionally — you won't just write code handed to you in a ticket. You're expected to understand the data you work on, contribute to requirements elicitation, own development and involving in all phases of testing, and take part in CI/CD and deployment processes. If you need a clear line between “analyst work” and “developer work”, this probably isn’t the right fit.

You know Snowflake well enough to choose the right tool for the job: when to write a SQL procedure, when Snowpark makes more sense, when a Dynamic Table is the answer, and when a simple view is enough. You don't reach for complexity when simplicity works.

We don't expect you to know everything from day one. You'll get full support from the team — domain knowledge, technical context, onboarding to our codebase and standards. There are people around you who will help you build internal knowledge, and we treat knowledge sharing as part of the job, not a side activity.

• Design and build transformation jobs in Snowflake — SQL stored procedures, Snowpark (Python/Scala), Dynamic Tables
• Engage in requirement analysis — understand the business and data context before writing code, not after
• Translate business and analytical requirements into clean, maintainable transformation logic
• Apply Snowflake best practices: clustering, micro‑partition awareness, query optimization, warehouse rightsizing
• Contribute to data modeling decisions — schema design, naming conventions, partitioning strategies
• Own CI/CD for your deliverables — branching, deployments, environment promotion via Bitbucket and Jenkins
• Write and maintain Airflow DAGs that orchestrate Snowflake transformation workloads
• Contribute to our data exchange layer (Database Roles, Secure Views, cross‑account sharing)
• Review code, pair with team members, contribute to team standards

• Must have:
  • 8+ years in data engineering with at least 2 years hands‑on Snowflake in production
  • Strong SQL — window functions, QUALIFY, MERGE, recursive CTEs, FLATTEN/LATERAL for semi‑structured data, performance tuning
  • Snowpark (Python or Scala) — building transformation logic, understanding when it’s preferable to pure SQL
  • Understanding of Snowflake execution model: micro‑partitions, clustering, pruning, query plan reading
  • Python or Scala for pipeline logic and tooling
  • Airflow — authoring DAGs, sensors, retries;
    Git, Bitbucket, CI/CD — comfortable owning deployments, not just writing code
  • Financial services or banking domain knowledge — you need to understand what the data means, not just move it. Credit risk, finance, or regulatory reporting background is expected
• Nice to have:
  • Snowflake Dynamic Tables — knows when to use them vs Tasks vs stored procedures
  • Streams and Tasks for incremental processing patterns
  • Snowpipe or Snowpipe Streaming
  • Data Stage — for working alongside legacy pipelines
  • Streamlit in Snowflake
  • SAFe or scaled agile experience

• Comfortable working across the full delivery lifecycle — analysis, development, testing, deployment
• Chooses the simplest Snowflake feature that solves the problem — not the most impressive one
• Reads the query profile before optimizing, doesn't guess
• Treats auditability, lineage, and…