STEM PhD AI Trainer - LATAM Remote
Listing checked September 18, 2026 · pay as published by Handshake AI
Overview
Handshake runs a year-round fellowship that puts STEM doctoral expertise to work on large language models. Researchers based anywhere in LATAM join a distributed group of subject-matter specialists who sharpen how AI systems handle advanced scientific and technical material. Assignments arrive asynchronously, so you set your own hours around a commitment of roughly 10 to 20 hours each week. Work centers on writing domain prompts and judging model responses, and you can spend part of the time researching topics that matter to your field with AI as a research partner. Project slots open by discipline, so timing depends on what is active when you apply.
What You'll Do8
- 1Write prompts that test how well a model reasons through problems in your scientific discipline
- 2Review model answers and score them against criteria you help define
- 3Document the reasoning behind each evaluation so other reviewers can follow your judgment
- 4Translate dense research concepts into plain task instructions a model can act on
- 5Research current topics in your field, using AI tools to speed up background reading
- 6Flag answers where the model misstates a core principle or invents a result
- 7Adapt your evaluation approach as project leads shift the focus of a study
- 8Track the hours you spend and report progress on assigned tasks
Requirements6
- 1Enrolled in or finished a doctoral program in Mathematics, Chemistry, Biology, Physics, or Engineering, or working in a postdoctoral post
- 2Reside in LATAM for the duration of the project
- 3Write clearly enough to turn complex research into structured instructions and scoring rubrics
- 4Confidence that your subject knowledge beats what current AI systems produce on key concepts in your field
- 5Comfort working on your own schedule without direct supervision
- 6Availability of about 5 to 20 hours per week while a project is active
Who Should Apply
Doctoral candidates, recent PhD graduates, and postdocs in math, chemistry, biology, physics, or engineering fit this best, mainly because the task asks you to judge whether a model got the science right. You should also be someone who enjoys writing. A large share of the work is turning your reasoning into instructions and criteria another person can read and reuse. The role suits people who want part-time, self-directed income alongside research or teaching, not a fixed 40-hour post. Candidates lose fit points when they submit a generic application that never names their subfield or a specific research area, since placement depends on matching your specialty to an open project. Weak writing samples also hurt. If your evaluations read as vague or your prompts lack precision, reviewers have little to judge your expertise by.
Salary Insight
Pay reaches up to $50.00 per hour, and you work as an independent contractor rather than an employee. That ceiling sits above typical rates for general data annotation and closer to what specialist scientific evaluation work pays, though the exact figure you receive depends on the project and your background. Hours are not guaranteed, so weekly earnings track how much time an active project gives you.
Pay and demand for Machine Learning & AI roles
AggregatedTypical pay
$75/hour
This role
up to $50/hr
Most Machine Learning & AI roles pay $55–$100 per hour. This role's pay sits below that range.
Based on 510 similar roles that publish pay · 90 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
- 567
- Listed in last 30 days
- 250
- Remote
- 97%
Hiring most right now: micro1 (259) · Mercor (79) · SME Careers (60)
Most requested skills · share of roles
- python17%
- technical writing9%
- llm evaluation8%
- ai evaluation7%
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
Up to $50/hr
Required Skills
Application Tip
Name your exact subfield and one concrete research area in your application, then attach or link a short writing sample that explains a hard concept from your field in a few clear sentences. Reviewers use that sample to judge whether you can build evaluation criteria, so precision matters more than length.
See NearSkill jobs more often in your search
Application & verification flow
1Instant rubric match
Your resume is scanned against this role’s requirements to check qualification fit.
2Screened before the employer sees it
Only profiles that clear screening are passed on.
3Outcome by email
We notify you at the address on your resume once the screening is reviewed.
Similar open positions
Explore active roles that match your skills and interests.
Handshake AI
VerifiedSTEM PhD AI Fellowship (Remote, Europe, UK)
Handshake AI runs a year-round fellowship that pulls STEM doctoral researchers into its AI research community, where experts shape how large language models handle scientific subjects. Fellows work remotely and set their own schedule, logging roughly 10 to 20 hours each week on project work. Typical tasks center on building domain-specific prompts and judging model responses for accuracy in mathematics, chemistry, physics, biology, and engineering. Project slots open and close with demand in each discipline, so placement depends on what is available when you apply.
Handshake AI
VerifiedSTEM Professor AI Fellowship - Remote Part-Time
Handshake AI seeks faculty in mathematics, chemistry, biology, physics, or engineering to join its AI research community as fellows. Fellows work from any location and set their own schedule, committing around 10 to 20 hours each week. The project centers on building domain prompts and judging how well large language models answer in your field. Assignments open throughout the year, though availability depends on the discipline. Project placement follows demand, so start dates vary.
Turing
VerifiedPh.D. / Postdoctoral / Master’s Expert
Turing, a San Francisco-based research accelerator, connects STEM experts with projects that fine-tune large language models such as ChatGPT. In this remote contractor role, you will design and solve challenging STEM problems to expose model weaknesses, then produce step-by-step solutions with well-articulated reasoning. You will collaborate with LLM researchers to align evaluation questions with areas where models struggle, including abstraction, multi-step reasoning, and symbolic manipulation. You may also help define new evaluation benchmarks drawn from physics curricula spanning early undergraduate to PhD-level topics.
Handshake AI
VerifiedComputer Science Expert - Remote AI Fellowship
Handshake AI runs a year-round fellowship that brings doctoral students and graduates in Computer Science together to improve Large Language Models. Fellows work from any location and set their own hours, with no minimum weekly commitment. Assignments focus on writing domain-specific prompts and reviewing model outputs for accuracy and reasoning. Project availability shifts by domain, so placement depends on current openings.
Handshake AI
VerifiedEngineering Expert - Remote AI Fellowship
Handshake AI brings together PhD and master's level engineers, with a focus on hardware engineering, to sharpen how large language models handle specialized technical subjects. The AI Fellowship runs all year, and project openings shift by domain and availability. Fellows work on a remote, asynchronous schedule for about 10 to 20 hours each week. Typical project tasks include writing domain-specific prompts and judging LLM responses, while also researching topics that matter to your field with AI tools at your side. The program accepts U.S.-based doctoral students, postdocs, and recent graduates who hold valid work or training authorization.
Handshake AI
VerifiedMath Expert AI Evaluation Fellowship Remote
Handshake AI recruits Math PhDs to sharpen how AI systems handle mathematical reasoning, proof construction, and technical problem-solving. The work centers on crafting difficult domain questions and checking AI-generated answers for accuracy, logical consistency, and mathematical rigor. Contributors join a remote, project-based fellowship with a schedule they set themselves. Most participants log 5 to 20 hours per week while a project is active, and no prior AI or technical experience is needed. Placement follows current project needs, with chances to join later projects as they open.