Experimental Biologist

DataAnnotation; United States; Sur place; USD 55 000 - 172 000; Temps partiel

Résumé du poste

AI Trainer Jobs is seeking experienced bench scientists to help train AI models by designing realistic experimental tasks from their own workflows. You will run tasks through frontier AI systems and grade the outputs against a professional standard, writing clear rubrics that specify necessary controls and evidence for pass/fail decisions.

Applicants should have 3+ years designing experiments in biotech or academic labs, with strong biology knowledge and the ability to explain data quality from

Qualifications

• 3+ years designing and running experiments in pharma, biotech, or academic labs.
• Hands-on practitioner who designs and troubleshoots experiments, not only follows SOPs.
• Clear written English and ability to judge data quality from raw outputs.

Responsabilités

• Design realistic experimental tasks from your workflows, including prompts and supporting files.
• Run tasks through frontier AI agents and grade their deliverables against professional standards.
• Write clear grading rubrics detailing controls, excluded wells, and failure diagnoses.
• Flag concrete failures with evidence from controls, normalization, and data.

Connaissances

Experiment design

Biology knowledge

Data analysis

Technical writing

Formation

Master’s or PhD in biology

Outils

Claude

ChatGPT

Description du poste

Pay: up to $40 – $125+ /hour (depends on task difficulty and specialization)

Location: Remote · Type: Independent contractor · Hours: Flexible hours · Payouts: Weekly

Overview

We’re looking for experienced bench scientists to help train AI models. You’ll design realistic experimental-biology tasks from your own practice, run them through frontier AI systems, and grade what comes back against the standard you’d hold a colleague to.

The models can already talk fluently about biology. What they can’t yet do reliably is the real work: call a plate QC go/no-go, gate a flow panel, check a cloning strategy, or read a Western blot from the raw images. Your bench judgment becomes the benchmark those models are measured against.

What you’ll actually do

• Design realistic experimental tasks from your own workflows: the scenario, the prompt, and the supporting files an agent would need (plate maps and reader output, gel and blot images, flow summaries, protocols, QC records).
• Run tasks through frontier AI agents and grade the deliverable they produce (a QC summary, experimental plan, or troubleshooting memo) against a professional standard.
• Write clear grading rubrics that specify what a correct deliverable must contain (the right controls, the right wells excluded, the right diagnosis of a failed run), and explain why a response passes or fails.
• Flag concrete failures with evidence: controls ignored, edge effects and outliers missed, wrong normalization, plausible-sounding troubleshooting the raw data contradict, or fabricated values.

Problems draw on whatever you know best:

• Molecular and cell biology: cloning, cell culture, and cell-based assays.
• Protein biochemistry: purification, characterization, and assay development.
• Immunology and flow cytometry: panel design and gating.
• Imaging and histology, microbiology and virology.
• In vivo and preclinical work, plus lab operations and analytical sciences.

Roles this fits

Common backgrounds: Research Scientist, Molecular / Cell Biologist, Postdoctoral Fellow.

What we look for

• 3+ years designing and running experiments in a pharma, biotech, or academic research lab (counted after undergraduate study).
• Depth in at least one of: molecular biology and cloning; cell culture and cell-based assays; protein biochemistry; immunology and flow cytometry; imaging and histology; microbiology or virology; in vivo and preclinical work.
• A hands-on practitioner who designs and troubleshoots experiments, not only executes SOPs: you can explain what each control is for and diagnose why a run failed.
• You generate and QC primary data, and can judge from raw instrument output whether a run is usable.
• Master’s or PhD (or current PhD candidate) in biology or a related field, completed in the U.S., Canada, Europe, or the UK.
• Clear written English, comfort with ambiguity, and familiarity with AI/LLM tools like Claude or ChatGPT.
• Based in the United States, Canada, or the UK (Ireland and Australia may also be accepted).

Compensation

Up to $40 – $125+/hr depending on task difficulty and specialization. Many contributors add $10k–$100k+ a year; some make it their full-time income.

About DataAnnotation

DataAnnotation is where 100k+ experts train the world’s leading AI models. $150M+ paid to contributors to date, and the average contributor stays 5+ years. Flexible, remote, and always project-available.