What the work actually looks like

You will read AI-generated clinical text — narrative assessments, summaries, structured flowsheet entries — and judge whether it matches what a pediatric inpatient nurse would actually chart. That means checking whether a respiratory assessment includes work-of-breathing descriptors and not just rate, whether pain documentation uses an age-appropriate scale, whether a neuro check on a post-op toddler is complete, and whether anything in the output would mislead a downstream clinician. A second stream of work is annotation: labeling and structuring real inpatient pediatric assessment data against a written guideline, applying the same rules across dozens of cases so the resulting dataset is internally consistent.

The hardest part is usually not clinical — it is guideline fidelity. Two nurses can chart the same shift differently and both be correct. Annotation projects require you to follow the project's definition even when your unit did it another way, and to flag the conflict rather than quietly substitute your own practice. Reviewers who raise clean, specific edge-case questions on a project channel tend to get invited back; reviewers who improvise do not.

What the screen looks for

  • Recency and setting. Bedside inpatient pediatrics within the past ten years, ideally a non-procedural med-surg or general peds unit rather than clinic, school, or purely procedural roles. Recent documentation experience outweighs total years licensed.
  • Concrete charting fluency. Expect to be asked what fields you completed on a shift assessment, what triggers escalation on your unit, and how you documented a specific scenario. Vague answers about "providing family-centered care" read as distance from the bedside.
  • Written precision. Much of the screen is typed. Feedback that says "this output is wrong" scores far below feedback that names the missing element and the clinical risk it creates.
  • Licensure. Active U.S. RN license, outside California.

Logistics

Fully remote, asynchronous, and contract-based. Most reviewers commit somewhere between 10 and 20 hours a week around clinical shifts, with work claimed from a queue rather than scheduled. Projects run in batches and can pause; treat this as supplemental rather than replacement income. You will need a reliable computer and internet, and comfort learning browser-based annotation tools quickly — the tooling changes between projects and there is rarely a long onboarding.