Data Engineer

About the role

ABOUT US

Most of the molecules driving human biology are invisible to us. Mass spectrometers already detect metabolites, lipids, and peptides, but the vast majority of those signals never get identified. A typical experiment names a small fraction of its features and discards the rest. We call this biology's dark matter. It’s signal-rich and mechanism-defining, yet almost entirely opaque.

Matterworks is building the foundation models that make that dark matter legible. Our Large Spectral Models do for biochemical biology what AlphaFold and ESM did for proteins: turning a library-bound discipline into something predictable and generative, and embedding it at every stage of R&D.

Come build the future of biological discovery with us.

Position Overview

Matterworks is seeking a Data Engineer to build and run the pipelines behind our models. Our platform acquires mass spectrometry and molecular data at scale and turns it into the datasets our AI team trains on and our product reasons over. You will own real pieces of that path end to end.

Our data serves two different customers. The AI team needs training corpora that are complete, correctly split, and reproducible. The product and agentic layer needs values a scientist can explain and we can safely show a customer. You will build inside the contracts and checks that keep both honest, and you will help extend them.

This is a hands-on role with a clear growth path. You will start by owning well-scoped pipelines and datasets and grow toward owning larger parts of the platform. You will report to the Head of Engineering and work daily with our machine learning researchers, scientists, and product team.

Key Responsibilities

- Build and Operate Pipelines: Own well-scoped pipelines end to end: designed, tested, instrumented, documented, and running on a schedule.

- Labels and Enrichment: Turn raw data into datasets people can actually use, with consistent schemas, trustworthy metadata, and documented definitions.

- Data Quality Checks: Design, build and extend our quality checks so that each build gets compared against the last one before it publishes, and a failure stops the pipeline instead of shipping.

- Ingest and Acquisition: Bring new public and partner datasets into the platform: fetching, converting, validating, and reconciling them against what we already hold. Expect messy scientific and vendor formats and file that require continuous improvements to our systems to handle at scale.

About You

- 2+ years of professional experience building data pipelines in production.

- Proficient in Python and SQL.

- Working knowledge of cloud data infrastructure. We run Argo Workflows and Metaflow on EKS, Glue and Athena over Apache Iceberg and Parquet, DuckDB, and Terraform. Depth in any comparable stack transfers fine.

- Demonstrated experience owning a pipeline or dataset end to end, including the tests, the monitoring, and the failures.