Drug Discovery Scientist

SaidGig; New York (NY); Remote; USD 83,000 - 124,000; Part time

Benefits offered by this job

Remote work

Hourly engagement

Claude Max subscription reimbursement

Job summary

SaidGig is seeking a researcher to design rigorous, realistic drug research challenges for frontier AI models. Challenges run in an offline sandbox with open-source tools and evidence sets that include incomplete and misleading information, requiring genuine scientific reasoning rather than pattern matching.

You will deliver a research prompt, a data room, and a grading document for each task, and construct defensible hidden conclusions with traceable evidence sources.

Qualifications

• Requires PhD or equivalent industry track record.
• Ability to formulate problems with rigorously defensible answers.

Responsibilities

• Create three deliverables for each task: a research prompt, a source-grounded data room, and a grading document.
• Design a challenge with a defensible hidden conclusion and the causal chain that makes that conclusion unavoidable.
• Assemble realistic evidence worlds using real data, ensuring every decisive value can be traced to a source.
• Write grading criteria that assess the scientific substance of a submitted memo.
• Complete verification and calibration before submitting each task.
• Own requested revisions and serve as the final authority on scientific accuracy and difficulty for assigned tasks.

Skills

Mechanistic enzymology

Structure-guided design

Fragment-based discovery

Structure-based medicinal chemistry

Education

PhD or equivalent

Job description

Role Overview

Design rigorous, realistic drug research and development challenges that evaluate frontier AI models. Each challenge places a model in an offline sandbox with open-source scientific tools and an evidence set containing incomplete, indirect, and occasionally misleading information. The correct conclusion is supported by the evidence but never stated outright, requiring genuine scientific reasoning rather than pattern matching or lookup.

Key Responsibilities

• Create three deliverables for each task: a research prompt, a source-grounded data room, and a grading document.
• Design a challenge with a defensible hidden conclusion and the causal chain that makes that conclusion unavoidable.
• Assemble realistic evidence worlds using real data, ensuring every decisive value can be traced to a source.
• Write grading criteria that assess the scientific substance of a submitted memo rather than its style or method.
• Complete verification and calibration before submitting each task.
• Own requested revisions and serve as the final authority on scientific accuracy and difficulty for assigned tasks.

Qualifications

• Working expertise in mechanistic enzymology, structure-guided peptide or antibody design, fragment-based discovery, or structure-based medicinal chemistry.
• PhD or equivalent industry track record.
• Ability to formulate problems with rigorously defensible answers that strong solvers cannot shortcut.
• Comfort evaluating the quality of scientific reasoning, not only final conclusions.

Work Terms

• Remote, hourly engagement.
• Commitment of at least 20 hours per week, with up to 40 hours available.
• Work is completed in a browser-based studio and Claude Code.
• A Claude Max subscription is required and fully reimbursed.

Compensation

• $60 to $90 per hour.