At Shopify, Staff Engineers use their expertise and passion to multiply the overall output of their team. As part of Observability within Infrastructure Engineering, you and the team own the systems that absorb and serve Shopify's telemetry — metrics, logs, traces, and more — at a volume that places us at the leading edge of the industry, ahead of companies substantially larger than us.
This is not a role about operating infrastructure tooling. We need a genuine software engineer with deep computing fundamentals. Our team is moving the core of its stack to Rust on thread-per-core, io_uring-based async runtimes, and we are building OLAP storage and query capability in-house rather than consuming it off the shelf. In the past year the team replaced a vendor metrics platform with a custom system written from scratch in roughly six months. That is the kind of engineering this role is for.
You will take ownership of an entire system or a broad swath of the OLAP surface — table and storage engine design, query execution, replication and coordination — with the latitude to challenge existing assumptions and rebuild from first principles where that is the right answer. The team is deliberately composed of senior, autonomous engineers running large projects independently, so you will have unusual scope and correspondingly high expectations.
What you'll be doing
- Designing and evolving OLAP storage and query internals: table and storage engine design, schema and partitioning strategy, materialized views, replication, and query optimization.
- Working in query execution and columnar storage — query planning, predicate and aggregate pushdown, columnar formats, statistics-based pruning.
- Writing systems-level Rust and Go against very large data volumes, with attention to memory movement, disk I/O, and per-core efficiency.
- Shaping distributed data architecture: separation of query and data nodes, scheduling and placement, metadata consistency, and replica coordination.
- Diagnosing and reasoning about live clusters at high volume, including greenfield R&D work where the problem is not yet well defined.
- Deep understanding of OLAP and columnar database internals — ClickHouse, ScyllaDB, DuckDB, or comparable systems.
- Fluency in query execution and columnar storage: query planning, predicate and aggregate pushdown, Parquet/Arrow, statistics-based pruning.
- Systems-level programming in Rust or Go. Bonus for async runtimes, io_uring or thread-per-core designs, NUMA awareness, and zero-copy techniques.
- Distributed data architecture experience: query/data node separation, scheduling and placement, metadata consistency, replica and consensus coordination.
- Demonstrable experience at high data volume, with the ability to explain the scaling failure modes you hit and how you addressed them.
- The ability to articulate in specific, nuanced detail the limitations and pitfalls of the systems and frameworks you have worked in — not just their strengths.
- Comfort with ambiguity: greenfield and R&D work, and the operability instincts to debug a live cluster.
- Effective use of AI tooling to accelerate your own work and cover gaps.
- While our stack includes ClickHouse, Rust, Go, Kubernetes, and Google Cloud Platform, experience with the exact components is not always required. Deep computing fundamentals and a willingness to learn matter more than any specific technology on the list.
Job description provided by Shopify.
View the original posting