Machine Learning Engineer

Mercor is hiring a Machine Learning Engineer to design, train, and deploy large-scale learning systems powering autonomous AI agents for its AI lab partner. This role sits at the intersection of ML research, systems engineering, and AI agent behavior, transforming ideas into robust, scalable learning pipelines. Responsibilities include designing and implementing scalable ML pipelines for training, evaluation, and continuous improvement; building and fine-tuning deep learning models for reasoning, code generation, and real-world decision-making; collaborating with data scientists to collect and preprocess training data; and developing benchmarking tools to test models across reasoning, accuracy, and speed. The engineer will implement reinforcement learning loops and self-improvement mechanisms for agent training, work with systems engineers to optimize inference speed, memory efficiency, and hardware utilization, maintain model reproducibility and version control with experiment tracking systems, and contribute to research efforts on learning strategies, fine-tuning methods, and generalization. The primary goal of this role is to develop, optimize, and deploy machine learning systems that enhance agent performance, learning efficiency, and adaptability, including model architectures, training workflows, and evaluation pipelines for autonomous intelligence and real-time reasoning. Work structure is as an hourly contractor to Mercor, paid weekly via Stripe Connect based on hours logged. The role is part-time (20–40 hours/week) with fully remote, asynchronous flexibility, allowing work from anywhere and on one’s own schedule. There is also a weekly bonus for tasks created. Mercor considers all qualified applicants without regard to legally protected characteristics and provides reasonable accommodations upon request.