ML Research Engineer - PhD - AI Trainer

mercor; United States; Remote; USD 103,000 - 124,000; Part time

Job summary

Mercor is seeking a Human Baseliner for Open-Ended ML Research Tasks on a contract basis. This remote role requires tackling open-ended ML problems within fixed time and compute budgets in a sandboxed Linux environment, using AI coding assistants as needed.

Ideal candidates have 3+ years of ML experience (PhD time counted), with a top-100 university background or FAANG experience, and expertise in PyTorch, JAX, or TensorFlow in at least one focus area such as pretraining or model architecture.

Qualifications

• 3+ years of ML experience (time in a PhD program counts).
• Attended a top-100 university or worked at FAANG or a comparable company.
• Experience with PyTorch, JAX, or TensorFlow.

Responsibilities

• Attempt open-ended machine learning research tasks under a fixed time and compute budget.
• Work independently in a sandboxed Linux environment with internet access.
• Use preferred tools, including IDEs and AI coding assistants like Cursor, Claude Code, and ChatGPT.
• Record full working sessions via screen recording.
• Complete pre-task and post-task questionnaires.
• Submit final work product, screen recording, and completed questionnaires for evaluation.

Education

Top-100 university

Tools

PyTorch

JAX

TensorFlow

Job description

Mercor connects elite creative and technical talent with leading AI research labs. Headquartered in San Francisco, our investors include Benchmark , General Catalyst , Peter Thiel , Adam D'Angelo , Larry Summers , and Jack Dorsey .

Position: Human Baseliner for Open-Ended ML Research Tasks

Type: Contract

Compensation: $75–$90/hour

Location: Remote

Commitment: 20+ hours/week

Role Responsibilities

• Attempt open-ended machine learning research tasks under a fixed time and compute budget.
• Work independently in a sandboxed Linux environment with internet access.
• Use preferred tools, including IDEs and AI coding assistants like Cursor , Claude Code , and ChatGPT .
• Record full working sessions via screen recording.
• Complete pre-task and post-task questionnaires.
• Submit final work product, screen recording, and completed questionnaires for evaluation.

Qualifications Must-Have

• 3+ years of machine learning experience . Time in a PhD program counts.
• Attended a top-100 university or worked at FAANG or a comparable company .
• Experience with PyTorch , JAX , or TensorFlow .
• Deep expertise in at least one focus area: pretraining, PPO , reward shaping, fine-tuning, LoRA , RLHF , architecture design, contrastive training, generative modeling, multilingual experience, or data pipelines.

Required Domain Expertise

• Practical experience in Pretraining , Reinforcement learning , Post-training , Dataset curation , or Model architecture .

Logistics

• One baseline attempt per contractor per task.
• Each task may only be attempted once.
• All work is confidential and covered by NDA.
• Compute and environment are provided; no personal GPU required.