Campus ML Research Engineer: Build Scalable AI in Finance

Tradermath; Greater London; Hybride; GBP 110 000 - 170 000; Plein temps

Résumé du poste

Jump Trading Group is seeking world-class engineers to join our research, trading, and engineering teams in building state-of-the-art ML systems for global financial markets. You will explore complex domains, collaborate with researchers and quants, and push the boundaries of AI from concept to production.

The role involves optimizing training pipelines on HPC clusters, integrating ML models where latency matters, and working with C/C++, Python, CUDA, and other low-level GPU languages in a

Qualifications

• Proficiency in Python and C++ for ML systems and production code.
• Experience with PyTorch, JAX, TensorFlow or similar DL libraries.
• Knowledge of GPU/accelerator programming (CUDA, Triton, SYCL, ROCm).
• Experience scaling ML systems across large datasets and HPC resources.
• Strong English communication and collaboration in teams.

Responsabilités

• Apply state-of-the-art ML techniques to complex domains in finance.
• Collaborate with researchers, quants, and engineers to build flexible ML frameworks.
• Optimize training pipelines to utilize HPC resources efficiently.
• Integrate ML models into latency-sensitive production systems.
• Work across C/C++, Python, CUDA and related low-level GPU languages.
• Build large-scale ML systems that are observable, performant, and maintainable.
• Perform other duties as needed.

Connaissances

Python

C++

PyTorch

JAX

TensorFlow

CUDA

GPU programming

English proficiency

Teamwork

Reliability

Outils

CUDA

Triton

SYCL

ROCm

Description du poste

Jump Trading Group is seeking world-class engineers to join our research, trading, and engineering teams in building state-of-the-art ML systems for global financial markets. You will explore complex domains, collaborate with researchers and quants, and push the boundaries of AI from concept to production.

The role involves optimizing training pipelines on HPC clusters, integrating ML models where latency matters, and working with C/C++, Python, CUDA, and other low-level GPU languages in a