Remote AI training and data labeling jobs

Umbrella listing for Mercor's energetic materials red-team panel: chemists and engineers or operators write benign, dual-use and adversarial prompts, judge frontier AI replies against a policy standard, and write reference answers. Remote contract at $65–75 per task; 16 hired this month.

Umbrella listing for Mercor's radiological safety red-team panel: RSOs, health physicists, source security, emergency response and nuclear medicine specialists write prompts, judge frontier AI replies against a policy standard, and write reference answers. Remote contract at $65–75 per task; 8 hired this month.

Evaluate AI finance outputs against rubrics, design hard finance tasks with worked solutions, and refine scoring guidelines for a leading AI lab. Needs 8+ years at a top-tier bank, asset manager or Big Four firm plus prior hands-on LLM rubric evaluation. US only, 35+ hours a week, $65–90/hour.

Certified pharmacy technicians answer medication questions and review AI responses for an AI lab building prior authorization workflows: dosing, interactions, indications, PA requirements. US only, 30–40 hours a week during the project. A flat $35/hour.

RF, microwave, antenna and electromagnetics engineers solve and critique hard technical problems for a short-term expert evaluation project: analysing systems and trade-offs, checking calculations and assumptions, and writing rigorous explanations. Hands-on industry experience and an EE-family degree. Remote hourly contract at $80–95/hour.

Build hard codebase-exploration tasks for an RL environment that trains AI agents: rewrite engineering questions about production Go repos (etcd, Traefik, Helm, gRPC-Go, Temporal and more) so frontier agents fail them, tune rubrics and foils, and pass a validation loop. $130 per approved task, remote.

Transcribe Italian audio, and English when required, edit transcripts for completeness, annotate data for AI training and flag unclear or low-quality recordings. Native Italian and professional transcription experience. 15 openings, contractor, $20–36/hour.

QA AI-agent runs inside the Financial Forecaster planning app: check the agent used the right scenario, account, coordinate and basis, catch plausible-but-wrong answers, harden tasks and sharpen grading. Needs weekly hands-on Financial Forecaster use and 5+ years in FP&A, reporting, technical accounting or lender reporting. $70–110/hour, remote.

Write or verify hard ten-option multiple-choice questions for an AI benchmark across clinical medicine, imaging, pharmacovigilance, health economics and rehabilitation, with step-by-step solutions and references. MD, DO, PhD or doctoral candidate, 10+ hours a week, asynchronous. $94–119/hour.

Write or verify 10-option multiple-choice benchmark questions in applied maths (signal processing, actuarial science, optimization, climate modeling and more), with chain-of-thought solutions and references. For maths PhDs and doctoral candidates. Remote, 10+ hours a week, $61–77/hour.

Author executable scientific-computing problems in ecology, biochemistry and genetics for Sci Code, a new AI benchmark: source a paper, dataset or repo, write the prompt and grading criteria, and keep it only if frontier models mostly fail. PhD plus Python or R, Git and Docker. 6 weeks, 20+ hours a week, $70/hour.

Senior materials scientists, and electrical or mechanical engineers, author realistic tasks with a prompt, a data room and a grading method, run them against an AI model and tighten them until the model can no longer reason through cleanly. Daily onboarding and office hours. Remote hourly contract at $60–90/hour.

Complete self-contained fund-ops exercises from mock ledgers, statements and notices: reconciliations, NAV variance attribution, trade-break investigations, corporate actions and payment exceptions, each graded against a rubric. 3+ years in fund admin or investment ops. US only, about 15 hours a week, $75–110/hour.

Full-time IB and M&A specialist embedded with a leading AI lab: vet model outputs on deal work, write instruction specs and golden solutions, and build finance benchmarks. 5+ years at a recognised institution, VP-level progression, MBA or CFA preferred. W-2 via Cincinnatus, hybrid Bay Area, $100–150/hour.

Remote hourly contract for developers on any stack who already use Visual Studio Code on their own Mac. $55–65/hour, paid weekly via Stripe or Wise. You need your own Mac and a display above 2.5 megapixels. Tasks are not described in the ad.

Grade AI-generated slides, spreadsheets and documents for real-world finance quality, catching factual, visual and presentation errors and writing structured feedback. Needs 5+ years at a top firm in the US, UK, Canada, Australia or New Zealand. $100–150/hour.

Build a real estate deal model in Excel from a synthetic document pack, with live formulas and sourced assumptions, then score two other contractors' models against fixed criteria. For associates at real-estate-focused mid-market PE funds with ~2 years of IB first. US only, about 20–25 hours over 1.5–2 weeks, $80–100/hour.

Full-time PE and VC specialist embedded with a leading AI lab: QA model outputs on investment work, write instruction specs and golden solutions, and design finance benchmarks. 5+ years at a recognised institution with Principal or VP-level ownership of decisions. W-2 via Cincinnatus, hybrid Bay Area, $110–150/hour.

Turn real lab and test-bench experience into AI benchmark tasks: build test logs, calibration records and failure reports, define the right diagnosis, and write 35+ point rubrics. Remote contractor, $30–70/hour paid per accepted task, 50 openings, 4+ years in test or production.

Annotate and score robotics video footage against a scoring guide: label actions, objects and events, rate clips for quality and relevance, and QA your own annotations. Mid-level annotation experience preferred, robotics footage experience heavily preferred. Flat $7/hour, 20 openings, remote contract.

Tag, segment and mark keyframes in video using Final Cut Pro on macOS to build AI training datasets. Professional or academic Final Cut experience counts; aimed at entry to mid-level editors. One opening, contractor, $15–80/hour.

Mechanical designers build and validate parametric FreeCAD models of parts and assemblies, applying GD&T, to create a CAD dataset for AI training. Optional Python scripting. Remote contractor role, 50 openings, $40–100/hour.

Turn everyday insurance judgment into AI training data: design realistic underwriting, claims, actuarial and compliance scenarios, review AI outputs and write feedback. Open to every insurance specialty, 3+ years, US-based. $75/hour.

Judge Vietnamese audio clips, from human speakers and AI speech models, for tones, pronunciation and naturalness, and back each rating with a written English explanation. Native Vietnamese and B2+ English preferred. 100 openings, contractor, $30–65/hour.