About
IMC is a research-driven trading firm that provides liquidity across trading venues and builds proprietary systems and algorithms for global markets. Its quantitative modeling, machine learning, and engineering support trading operations and value and risk management for investors.
Responsibilities
•Develop large-scale distributed training pipelines for datasets and complex models
•Build and optimize low-latency inference pipelines for real-time predictions in production
•Develop libraries to improve machine learning framework performance
•Maximize training and inference performance using GPU hardware and acceleration libraries
•Design scalable model frameworks for high-volume trading data and real-time, high-accuracy predictions
•Collaborate with quantitative researchers to automate ML experiments, hyperparameter tuning, and model retraining
•Partner with HPC specialists to optimize workflows, improve training speed, and reduce costs
•Evaluate and roll out third-party tools for model development, training, and inference
•Extend open-source ML tools and improve their performance
Requirements
•5+ years of experience in machine learning focused on training or inference systems
•Strong engineering skills, including Python, CUDA, or C++
•Knowledge of machine learning frameworks such as PyTorch, TensorFlow, or JAX
•Proficiency in GPU programming for training and inference acceleration, e.g. CuDNN or TensorRT
•Experience with distributed training for scaling ML workloads, e.g. Horovod or NCCL
•Exposure to cloud platforms and orchestration tools
Nice to have
•Hands-on experience with real-time, low-latency ML pipelines in high-performance environments
•Contributions to open-source projects in machine learning, data science, or distributed systems
