Machine Learning Research Expert
Listing checked October 8, 2026 · pay as published by AfterQuery
Overview
AfterQuery supplies the expert evaluation data that frontier AI labs rely on to benchmark and refine their models. Researchers here write ML research problems pulled from their own work, then solve each one to a reference answer covering experiment design, ablation studies, model evaluation and failure analysis. The work runs remote and async, with no fixed schedule, at $115 per hour. Judgment carries more weight than volume, and the problems you write shape how the next generation of AI systems handles machine learning research.
What You'll Do4
- 1Author research problems that reflect your own specialty, spanning experiment design, ablations, evaluation and how models behave in practice.
- 2Carry each problem through to a worked reference answer, showing the reasoning that connects every step.
- 3Flag the subtle research mistakes that read as plausible but only a practitioner would catch.
- 4Write up each solution in clear English so another researcher can follow the path from problem to answer.
Requirements5
- 1Master's or PhD in computer science, statistics, mathematics or a closely related discipline.
- 2At least 3 years of full-time ML research after finishing your degree. Postdoc time counts; years spent inside a PhD program do not.
- 3Direct experience designing experiments, running ablations or evaluating models.
- 4Strong Python plus a deep learning framework such as PyTorch or JAX.
- 5Availability of 10 hours or more each week.
Who Should Apply
The right candidate holds a graduate degree in CS, statistics or math and has spent at least three years past graduation doing ML research, in a university group or an industry lab. You design experiments, run ablations, evaluate models, and can walk a reader through the reasoning behind a result. Publications at NeurIPS, ICML, ICLR, ACL or CVPR strengthen an application. The role fits researchers rather than ML engineers whose work centers on deployment, data engineers or data annotators. Applications lose ground when candidates count PhD years toward the three-year research requirement, which the listing rules out, or submit problems with thin evaluation sections that skip failure analysis.
Salary Insight
Pay is set at $115 per hour. That rate sits near the top of contract work for expert evaluation tasks and usually matches researchers who hold a completed graduate degree and several years of independent ML research behind them. Payment arrives weekly through Stripe, and you decide how many hours you take on each week.
Pay and demand for Machine Learning & AI roles
AggregatedTypical pay
$70/hour
This role
$115/hr
Most Machine Learning & AI roles pay $40–$98 per hour. This role's pay sits above that range.
Based on 1,035 similar roles that publish pay · 302 publish only a top rate; those count at the rate they gave
Rates shown per hour. Yearly and monthly pay converted; one-time fees and non-USD pay are not included.
- Live similar roles
- 1,182
- Listed in last 30 days
- 440
- Remote
- 98%
Hiring most right now: micro1 (342) · SME Careers (253) · Mercor (144)
Most requested skills · share of roles
- ai training14%
- llm evaluation12%
- ai evaluation12%
- data annotation9%
Figures from Machine Learning & AI roles live on NearSkill when this page loaded. A role can close before you apply, so check the listing itself.
Compare your resume against these rolesLocation
Compensation
$115/hr
Required Skills
Application Tip
Lead with the details of a research project you ran: the hypothesis, the ablation you designed, and the result that surprised you. Name the publication venue if you have one, point to work you shipped in PyTorch or JAX, and state that you can give 10 hours or more each week. Evidence of evaluation and failure analysis separates research judgment from engineering output, which this role does not want.
How NearSkill helps you apply
1. Check your job match
We compare your resume with the job requirements to help you understand how well your experience fits.
2. Continue to the employer’s website
We’ll direct you to the employer’s application portal, where you can complete the application.
3. Get updates by email
If NearSkill is handling follow-up for this role, we’ll email you when there’s an update on your screening.
Want to check your fit first?
See how your resume matches this role before applying.
Check my job matchSimilar open positions
Explore active roles that match your skills and interests.
Mercor
VerifiedMachine Learning Expert (Remote, Part-Time)
Mercor runs a standing pool for machine learning practitioners who provide the expert judgment that frontier AI research teams cannot produce on their own. Your standards become the rubric that grades model outputs. Projects are remote and part-time, and clients set rates from $70 to $120 per hour based on scope and depth. Work can involve writing a modeling problem from a system you shipped, checking training and evaluation code that models produce for loss-objective alignment and eval leakage, or catching an approach that would train without error but fail in production because of distribution shift, label noise, or a metric that rewards the wrong behavior.
Handshake AI
VerifiedMachine Learning Researcher - AI
Handshake AI hires computer science and AI/ML researchers for hourly, project-based work that improves how models handle technical research. You will judge AI-generated answers, build expert training examples, and give written feedback grounded in machine learning, artificial intelligence, and CS research methods. The role is remote and asynchronous, with no minimum hours; contributors on active projects tend to log about 5 to 20 hours each week. Placement hinges on current project needs, and strong contributors may qualify for later projects. The listing notes you do not need prior experience with AI training work.
AfterQuery
VerifiedML Research Expert - Benchmark Authoring
AfterQuery seeks Master's and PhD-level ML researchers to design and validate benchmark tasks that measure what frontier models can do across machine learning. The work is remote, asynchronous, and flexible, with 5 to 40 hours a week available. Your domain expertise shapes the tasks, whether you focus on language models, reinforcement learning, robotics, ML systems, or another listed area. The role fits people who write ML code and publish results, not those with annotation or data-labelling backgrounds. Pay is $150 per hour.
AfterQuery
VerifiedMachine Learning Data Science Expert
AfterQuery, an early-stage Y Combinator-backed company, seeks remote machine learning and data science experts for contract projects. The work matches tasks common to data science, ML engineering, AI research, or analytics roles, and it runs over the next 2 to 3 weeks on your own schedule. Projects arrive one at a time, with a listed baseline of 10 hours a week and expected commitments of 10 to 20 hours per week for accepted work. You will build and evaluate models, develop predictive models, and tune hyperparameter settings while documenting your methods. Pay starts at $60 per hour and can reach $100 per hour, and the project format helps you add real-world ML work to your portfolio.
AfterQuery
VerifiedScientific Computing Research Expert
AfterQuery builds a network of PhD-level and PhD-pursuing researchers who write and calibrate evaluation tasks for AI models. Your starting point is real computational work you have run, debugged, or been paid to do, such as data reduction, model fitting, simulation setup, instrument data processing, numerical solvers, or inverse problems. Tasks span life sciences, physical sciences, earth sciences, mathematical sciences, and engineering sciences. Difficulty comes from the science itself, not from withheld information. You can take on tasks as an independent contractor, with Python as the core tool for running and debugging your own work.
AfterQuery
VerifiedData Scientist Expert
AfterQuery assembles the expert evaluation data that frontier AI labs rely on to test and sharpen their models. This contract role puts practicing data scientists to work authoring the analysis problems those evaluations run on, pulled straight from your own daily work in experiments, metrics, forecasting and predictive modeling. You write each problem, solve it end to end with a reference answer, and flag the mistakes that pass a quick look but fail under a practitioner's eye. Work is fully remote and async: no fixed hours, and you decide how much or how little you take on each week.