ML Engineer: Large-Scale Training & Real-Time Inference

IMC Financial Markets; New York (NY); On-site; USD 175,000 - 250,000; Full time

Benefits offered by this job

Discretionary bonus

Paid leave

Insurance

Job summary

IMC Financial Markets in New York, NY is hiring a full-time Machine Learning Engineer to build scalable systems for training and deploying large-scale ML models across the firm. The role combines ML Research Engineering and ML Performance Engineering to accelerate experiments and deliver efficient production execution for trading.

You will develop distributed training pipelines, partner with quantitative researchers, and optimize HPC workflows and GPU-accelerated inference to deliver real-time

Qualifications

• 3+ years of experience in machine learning focused on training or inference systems.
• Production software engineering in Python with testing discipline.
• Strong ML fundamentals: reasoning about model behavior and performance.

Responsibilities

• Develop large-scale distributed training pipelines for complex datasets and models.
• Collaborate with quantitative researchers to automate ML experiments and retraining.
• Work with HPC specialists to optimize workflows and reduce compute cost.
• Build and optimize low-latency inference pipelines for real-time predictions.
• Maximize training and inference performance using GPU hardware and acceleration libraries.
• Extend internal ML frameworks and open-source tools to boost performance.

Skills

ML experience

Python

ML fundamentals

Systems mindset

Collaboration

Job description

IMC Financial Markets in New York, NY is hiring a full-time Machine Learning Engineer to build scalable systems for training and deploying large-scale ML models across the firm. The role combines ML Research Engineering and ML Performance Engineering to accelerate experiments and deliver efficient production execution for trading.

You will develop distributed training pipelines, partner with quantitative researchers, and optimize HPC workflows and GPU-accelerated inference to deliver real-time