Senior Research Engineer (ML infrastructure )

We are looking for a Senior Research Engineer to build and improve the ML infrastructure that powers our deep learning research and trading.

Our DL strategies are becoming an increasingly important driver of the business, and the infrastructure behind them has a direct impact on how quickly researchers can test ideas, how efficiently we use compute, and how reliably successful research reaches production across markets and asset classes.

You will work at the boundary between deep learning research and infrastructure. Your job is not just to maintain ML systems, but to identify what is slowing research down, solve technically difficult problems across the stack, and turn research prototypes into reliable, scalable production systems.

The scope is broad: training frameworks, datasets and data pipelines, distributed training, experiment infrastructure, reproducibility, observability, and tooling around the research lifecycle. Depending on your strengths, you may go deep in one of these areas or work across several of them.

This is a highly hands-on IC role with substantial technical ownership. You will work closely with researchers, other research engineers, and infrastructure teams, and will be expected to independently drive important problems from diagnosis to production.

• Turn successful research prototypes into robust implementations that can be trained, validated, and deployed across multiple markets and asset classes;
• Work directly with researchers to remove infrastructure bottlenecks from the research loop — make new ideas easier to prototype, experiments faster to run, and failures easier to understand;
• Build and improve our internal ML platform — training framework, datasets and data pipelines, orchestration, experiment tracking and reproducibility tooling, GPU/compute infrastructure, and tooling for releasing models into production;
• Own technically challenging areas of the ML stack end-to-end: identify problems, design solutions, implement them, measure their impact, and maintain them in production;
• Make training reproducible and observable: investigate regressions, debug models that no longer reproduce from master, and improve tooling around data, experiments, model quality, and training behavior;
• Take ownership of shared training, dataset, and research infrastructure code — proactively find bugs, reduce technical debt, improve abstractions, and maintain a high bar through rigorous code review;
• Work across team boundaries when a research problem spans datasets, storage, compute, training infrastructure, or production systems;
• Identify areas where researchers are repeatedly paying an infrastructure tax and build reusable solutions instead of fixing the same problem case by case;
• When useful, contribute directly to research: run experiments, investigate model behavior, prototype architectural or optimization ideas, and help push model quality forward.

• A strong, versatile ML systems / research engineer who is comfortable working on ambiguous problems at the intersection of deep learning and infrastructure;
• Strong background in at least one of:
• large-scale ML training systems and distributed training;
• ML/data infrastructure, including datasets, storage formats, orchestration, experiment tracking, and configuration;
• research engineering around large-scale deep learning systems;
• Hands-on experience making ML research faster or more reliable: for example, by reducing training time, improving hardware utilization, accelerating data access, improving experimentation workflows, or eliminating recurring infrastructure bottlenecks;
• Ability to independently investigate complex systems: profile them, form hypotheses, run experiments, trace problems across abstraction boundaries, and arrive at practical solutions;
• Strong engineering judgment and a high bar for code quality, reproducibility, observability, and maintainability in a large shared ML codebase;
• Strong product sense toward internal research infrastructure: you care not only whether a system works, but whether researchers can use it effectively and iterate quickly;
• Ability to work closely with researchers, understand what they are trying to achieve, and translate research needs into infrastructure improvements that compound over time;
• Strong ownership and initiative: you notice important problems that do not have a clear owner and are willing to take them from an unclear state to a working solution;
• Ability to influence technical direction through expertise, design work, code review, and collaboration without requiring formal management authority;
• While prior trading experience can be useful, it is not a prerequisite. Our priority is a first-principles mindset, strong technical depth, and a willingness to rethink how quantitative ML research infrastructure should work.

• High base salary and social benefits;
• Generous bonus structure. We are very flexible in discussing salary and conditions of employment;
• Cutting-edge hardware and software in production, as well as high technical expertise across the company, which allows us to implement bold ideas and achieve great results. Ownership over initiatives that directly solve business problems;
• Ability to trade on dozens of international exchanges;
• Flexible workflow (lack of formalism and bureaucracy, no pressure and over-management) and working schedule;
• Tuition reimbursement, conference and training sponsorship.