Senior Staff Software Engineer - Data Platform
About the role
What you'd own
The lakehouse and the streaming ingestion that feeds it. Iceberg/Delta on S3, Spark/Glue processing, Kafka and CDC pipelines, the abstractions that let other teams land data, build pipelines, and publish datasets without inventing the patterns themselves. You’d help set the multi-year technical direction, design the golden paths, and ship the work — directly, with your hands on the code that matters.
How the team works
You’d report to the Director of Data Engineering and partner with a peer Manager whose scope is people leadership and operational delivery. You own technical direction, application architecture, and the engineering bar. You’ll work alongside other senior staff engineers across the broader data platform org, and partner with security, compliance, and the consumer teams who depend on what we ship.
What "hands-on" actually means here
Worth being specific, because “hands-on” means different things to different people:
- Yes: writing production code on the platform's hardest problems, owning the design and build of new abstractions end-to-end, leading design reviews, setting the engineering bar through the work itself.
- Sometimes: prototyping, deep diving on a tricky migration, pairing with engineers on tough problems.
What the work looks like
- Modernization. Building out the lakehouse — the patterns Marqeta will run on for years. New abstractions, new tooling, the paved road internal teams want to use.
- Live system work. Evolving and hardening the platform that's already in production. Reliability, performance, developer experience — improved without disrupting the teams who depend on it.
- Engineering excellence. Testing confidence, release velocity. On call, blameless postmortems.
What you bring
We care more about demonstrated work than years on paper, but the experience this role asks for typically takes around a decade to accumulate.
- A career building data platforms, not just pipelines. You can name the abstractions you've designed, the teams that adopted them, and the problems they solved. If most of your last several years has been writing transformations rather than building the system transformations run on, this isn't the right shape.
- Production depth in modern lakehouse work. Iceberg or Delta in production. Strong fluency in the surrounding stack: AWS data services (S3, Glue, EMR), Spark, Airflow, Kafka, IAM boundaries for multi-tenant data access. You don't need depth in all of it — you need real depth in the lakehouse and credible engagement with the rest.
- Mastery of Python in a production data-platform context, with credible depth in Go or Java.