Talk about your next project
PostHog case study

The lesson

Show the automation claim through the human review loop and the small product fixes it creates.

The move

The story opens inside PostHog's customer-story interface, with a Juno hero card, the title, date, company description, tools used, industries and users. The first section explains Juno's health tracking problem, its earlier agent system pulling from Mixpanel, the switch to PostHog, and a quote saying Juno's merge rate for self-driving PRs is roughly 90%. A section called "Self-driving development and user privacy" explains how Juno strips PII and PHI, then shows specific fixes including "Log all" for medication logging, an unresponsive Continue button and a question flow that needed a clear selection state and a Next button. A later section, "Are self-driving PRs perfect?", keeps the claim grounded by saying Marshall still reads every PR.

Why it works

  • Our readThe tools, industry and user labels let a reader place Juno before getting into the story.
  • Our readThe story makes the AI claim concrete by naming the workflow: spot the issue, open a PR, review the code and merge.
  • Our readThe privacy section answers the obvious healthtech objection before showing the product fixes.

What not to copy

Don't copy Juno's chronic illness context, medication examples or self-driving PR claim. Keep the structure that moves from customer pressure, to the old workaround, to the new workflow, to specific fixes and a candid limit.

More from PostHog

More case studies

Steal it for your brand.

Tell us what you want to make. We’ll talk through how to put these ideas to work for your brand.