Case Study
KJ Technology, internal tooling.
Turning a completed client automation into reusable case study content required a dedicated writer's time — gathering the facts, drafting three different versions (a tight proof point for proposals, a narrative case history for sales conversations, and a full one-page write-up for the website), and manually tracking review status. That write-up step routinely became the bottleneck; good projects went undocumented because nobody had a day free to write them up.
A structured intake form captures the seven facts that make a case study credible: client context, the before-state, the problem it created, what was built, what made the solution thoughtful, what changed, and the proof. An AI skill turns those answers into all three formats at once — in KJ's voice, tagging every metric as verified or estimate rather than blending the two. The output lands directly in a case study library with Draft, Approved, and Published stages. A case study only goes live after a person actually reviews it — the AI drafts, it doesn't approve itself.
Case study turnaround dropped from a multi-day writing task to a same-day draft ready for review. Every case study still passes through the same human approval gate before publishing.
Verified: this exact workflow produced the other entries in this site's Proof section — it's proof of itself.