Role Overview
Machine Learning Engineers at IMC build systems for training and deploying large-scale ML models across global operations. They collaborate with quantitative researchers, hardware specialists, and software engineers to build robust systems leveraging GPU acceleration, distributed computing, and open-source ML ecosystems. The role spans ML Research Engineering (distributed training pipelines, experiment automation) and ML Performance Engineering (low-latency inference, GPU optimization), with overlap between areas.
Responsibilities
- Develop large-scale distributed training pipelines managing complex datasets and models.
- Partner with quantitative researchers to automate ML experiments, hyperparameter tuning, and model retraining.
- Work with HPC specialists to optimize workflows, improve training speed, and reduce compute cost.
- Build and optimize low-latency inference pipelines for real-time predictions in production.
- Maximize performance in training and inference using GPU hardware and acceleration libraries.
- Extend and improve internals of ML frameworks and open-source tools for workload performance.
- Design scalable model infrastructure for high-volume trading data and real-time predictions.
- Evaluate and roll out third-party tools enhancing model development, training, and inference.
Requirements
- 3+ years experience in machine learning with focus on training or inference systems.
- Production software engineering: Python, testing discipline, ownership of shipped systems.
- Machine learning fundamentals: understand modern ML workloads beyond high-level APIs.
- Systems mindset: measurement-driven decisions on performance, scale, reliability; sound build-versus-buy decisions.
- Strong collaboration: explain technical decisions, work through differing perspectives, adapt based on evidence.
- Depth in one or more: distributed training (PyTorch, TensorFlow, JAX), GPU performance (CUDA, Triton), production inference systems, numerics and correctness at scale.
Nice to Have
- Exposure to cloud platforms and orchestration tools.
- Track record of contributing to open-source projects in ML, data science, or distributed systems.
About Us
IMC is a research-driven trading firm where quantitative modeling, ML, and engineering shape modern markets. A stabilizing force since 1989, providing liquidity across global venues using proprietary technology and capital.