Research Engineer, Pre-Training

Trading Interview; New York (NY); On-site; USD 300,000 - 350,000; Full time

Benefits offered by this job

Medical insurance

Dental insurance

Vision insurance

Employer match 401(k)

Paid parental leave

Wellness programs

Job summary

Jump Trading Group is seeking a Pre-Training Research Engineer to lead end-to-end training of massive foundation models. You will build scalable infrastructure across thousands of GPUs/TPUs, stream terabytes of data, and design efficient kernels for 10x gains.

You will collaborate with researchers to push boundaries in mixed-precision training and model parallelism, delivering impactful results for live trading.

Qualifications

• Experience with large-scale distributed training and performance optimization.
• Publications in efficient training methods or ML systems are a plus.
• Strong background in numerical computing, HPC, or distributed systems.

Responsibilities

• Own the training stack for massive foundation models across GPUs/TPUs.
• Build fault-tolerant infrastructure and data pipelines processing terabytes per second.
• Co-design architectures with researchers and optimize training performance.

Skills

Distributed training

Research publication

Python

PyTorch

JAX

Communication skills

Reliability

Education

MS or PhD in CS/ML/Physics/Math

Tools

Kubernetes

Slurm

NVLink/InfiniBand

OS internals

Job description

Jump Trading Group is committed to world-class research. We empower exceptional talents in Mathematics, Physics, and Computer Science to seek scientific boundaries, push through them, and apply cutting-edge research to global financial markets. Our culture is unique. Constant innovation requires fearlessness, creativity, intellectual honesty, and a relentless competitive streak. We believe in winning together and unlocking unique individual talent by incentivizing collaboration and mutual respect. At Jump, research outcomes drive more than superior risk-adjusted returns. We design, develop, and deploy technologies that change our world, fund start-ups across industries, and partner with leading global research organizations and universities to solve problems.

Our team is a group of quantitative researchers, engineers, and ML experts leading foundation model research and trading at Jump. Our mission is to combine emerging techniques and original research to generate signals from financial market data and monetize it globally. We are building the future of ML-powered trading through breakthrough foundation models, and we’re looking for an exceptional Pre-Training Engineer to join our team.

What You'll Do:

As a Pre-Training Research Engineer, you’ll be at the forefront of developing massive-scale foundation models that fundamentally transform how we understand and predict markets. You’ll own and drive the entire training stack: building fault-tolerant infrastructure that scales across thousands of GPUs and TPUs with near-linear performance, engineering data pipelines that stream terabytes per second as our models train on petabytes of data from every corner of the global markets, and designing custom kernels that unlock 10x efficiency gains. Co-designing novel architectures with researchers and pioneering cutting-edge approaches to mixed-precision training and model parallelism, you’ll have the latest generation hardware at your disposal. This isn’t incremental optimization; we’re pushing the boundaries of what’s possible in pre-training at scale, where your improvements directly impact live trading.

Other duties as assigned or needed.

Skills You'll Need:

• Expertise and track record of significant, measurable performance improvements in large-scale distributed training (MFU, throughput, convergence, cost-per-token).
• Published research in efficient training methods, scaling laws, architectures, or systems for ML
• Background in numerical computing, HPC, or distributed systems, including familiarity with GPUs/TPUs, high-performance networking (NVLink/InfiniBand), Kubernetes/Slurm, and OS internals
• Expertise in Python and deep experience with modern deep learning frameworks (PyTorch and/or JAX)
• Advanced degree (MS or PhD) in Computer Science, Machine Learning, Physics, Mathematics, or a related quantitative field, or equivalent industry experience at a frontier lab
• Ability to balance ambitious research goals with practical engineering constraints
• Strong problem-solving skills, results orientation, and excellent collaborative communication
• Reliable and predictable availability

Bonus Points:

• Expertise in: CUDA kernel development, Triton/Pallas/CuTe DSLs, PyTorch/JAX internals, XLA optimization, or hardware acceleration (FPGA/ASIC)
• Knowledge of reinforcement learning, post-training, or fine-tuning techniques
• Knowledge of financial markets or trading
• Discretionary bonus eligibility
• Medical, dental, and vision insurance
• HSA, FSA, and Dependent Care options
• Employer Paid Group Term Life and AD&D Insurance
• Voluntary Life & AD&D insurance
• Retirement plan with employer match
• Paid parental leaveWellness Programs

Annual Base Salary Range

$300,000 — $350,000 USD

Jump Trading Group is committed to world class research. We empower exceptional talents in Mathematics, Physics, and Computer Science to seek scientific boundaries, push through them, and apply cutting edge…