At our October Healthcare AI Build Club session, two members showed AI tools they built that rework existing content without adding anything new.

The Healthcare AI Build Club is a monthly Zoom call, hosted by Dawn Halpin, where people in healthcare show something they built with AI and talk through what worked. There are no slide decks.

There is also a Slack group for ongoing discussion in between sessions. Since its first session in September, the club has grown from about 20 members to more than 100.

Our takeaways

  • Sarah Cohan showed a system she built in Claude that rewrites marketing copy for a new audience, such as health plan messaging rewritten for employer benefits leaders.
    • A list of locked terms keeps product names and regulated outcome claims from being reworded or made up.
    • A dashboard flags changes for human approval and runs compliance checks.
  • Tamar Rucham showed custom patient reports for a longevity and preventive care platform, built in each clinic's own style, structure, and voice.
    • A second AI model audits each report against the source data and corrects any mismatch.
    • Clinicians review every report.
  • Tamar's rule that the AI adds no new content echoed Sarah's approach: in both builds, the AI reshapes what's already there and something else checks the result.

If you're building with AI in healthcare and want to join the club, email Dawn at dawn@paubox.com.

See related: The 2026 Benchmark for Building with AI in Healthcare