[LONG PROJECT] Anonymization Writer

About the Role

We're hiring Writers on a contract basis for Project Sisyphus, a PII anonymization program. You work through business documents and structured records (CRM exports, email, tickets, chat logs, accounting data), deciding what is and isn't personal information, catching anything that was missed, and grouping related entities so the same person or organization is handled consistently. Reviewers check your work before it ships.

We are working to build a lasting team. This is an internal project that will run long into the future, and we want writers we can build lasting teams and structures around. Strong writers are the pool we promote reviewers from.

What You'll Do

• Confirm, reject, add and correctly type PII across tasks that contain dozens to hundreds of individual decisions

• Group related entities (people, organizations, contact details) into the right clusters

• Find personal information that automated detection missed, and avoid tagging things that aren't personal

• Apply guidelines that are updated several times a week, using the version in force on the day you work

• Act on reviewer feedback quickly, and apply it across your other tasks rather than fixing one instance

• Work asynchronously with a distributed team, and raise ambiguous cases in the project channel rather than guessing

What We're Looking For

• Able to commit 30+ hours per week, with this as your main project

• Prior experience with AI data work (labelling, rating AI responses, or QA/review of AI outputs)

• A sharp eye for detail across long, repetitive material, and the patience to double-check your own work

• A track record in detail-heavy work such as operations, accounting, editing, design, compliance or account management

• Comfort with rules that change frequently, and a habit of checking the current guideline before relying on memory

• Takes feedback well and adjusts quickly

• Ability to work entirely without AI assistance on task data

• Excellent written English; reliable internet and a quiet, focused workspace (US-based preferred)

Nice to Have

• Familiarity with PII, privacy or data-protection concepts (what counts as an identifier, direct versus indirect identification)

• Experience with medical record de-identification, copy-editing or paralegal work

• Experience reading structured business data such as CRM, ticketing, email or accounting exports

We consider all qualified applicants without regard to legally protected characteristics and provide reasonable accommodations upon request.