The Research Engineering team develops and deploys the pipelines connecting a research idea all the way to its rollout in production. We exist to ensure that the entire pipeline describing our research is efficient, effective, and scalable. People in our team are avid collaborators and work in tight feedback loops with our researchers. We talk through problems on desk, prepare solutions same day, and work out how to deploy safely for tomorrow.

The scope of the team covers anything that accelerates the path from research to production at massive scale. This could be designing our external data ingestion pipelines, writing our feature discovery and storage systems, building our model validators and experiment trackers, authoring and deploying a bespoke algorithm, or figuring out how to roll out something novel.

We use a broad toolkit including Python, C++ and internal domain-specific languages, and are happy to pick up new tools and languages as required.

About you

• You write production code in both Python and C++. You reach for the second one when it matters, and you know when it does.
• You're naturally curious about how other teams' systems work, not just your own, and you often end up being the person who spots where things won't fit together before it becomes a problem.
• You enjoy working on many projects at once, seeing connections and opportunities between them which others miss.
• You’re motivated to solve issues with whatever tool is best, rather than a preference towards deep specialisation in one area.
• You like working with research stakeholders across the business and you're good at figuring out what someone actually needs, not just what they asked for.
• You've done this commercially. You've owned something in production that other people relied on. Not necessarily big, but something you had to maintain, debug and live with long after you shipped it

What you don’t need

A finance background. Every language we touch. We hire engineers who like to solve the problems our traders face - we'll teach you the context on the ground.