Apply on Employer Site

hackajob · 2 days ago

Machine Learning Engineer - India (Train AI Models Part Time!)

United States

Full-time, Part-time

Remote

Mid Level

$500/wk - $1,000/wk

hackajob is collaborating with Mercor to connect them with exceptional tech professionals for this role. Mercor is hiring a Machine Learning Engineer to help design, train, and deploy large-scale learning systems powering autonomous AI agents for its AI lab partner.

Artificial Intelligence (AI)Generative AIHuman ResourcesRecruitingSoftware

Responsibilities

Design and implement scalable ML pipelines for model training, evaluation, and continuous improvement

Build and fine-tune deep learning models for reasoning, code generation, and real-world decision-making

Collaborate with data scientists to collect and preprocess training data, ensuring quality and representativeness

Develop benchmarking tools that test models across reasoning, accuracy, and speed dimensions

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, integrating with experiment tracking systems

Contribute to cross-functional research efforts to improve learning strategies, fine-tuning methods, and generalization performance

Qualification

Machine LearningDeep LearningPythonMLOps toolsPyTorchTensorFlowReinforcement LearningTraining InfrastructureEnd-to-end ML systemsAI AgentsModel PerformanceData-driven ExperimentationCritical ThinkingCollaboration

Required

Have a strong background in machine learning, deep learning, or reinforcement learning

Are proficient in Python and familiar with frameworks such as PyTorch, TensorFlow, or JAX

Understand training infrastructure, including distributed training, GPUs/TPUs, and data pipeline optimization

Can implement end-to-end ML systems, from preprocessing and feature extraction to training, evaluation, and deployment

Are comfortable with MLOps tools (e.g., Weights & Biases, MLflow, Docker, Kubernetes, or Airflow)

Have experience designing custom architectures or adapting LLMs, diffusion models, or transformer-based systems

Think critically about model performance, generalization, and bias, and can measure results through data-driven experimentation

Are curious about AI agents and how models can simulate human-like reasoning, problem-solving, and collaboration