Clayton Rodgers
← All case studies File No. 01

Conversational AI Assessment Agent

Status: Shipped For: Dentsu International Built with: Claude, TODO: add stack

The problem

TODO: What was broken about the existing assessment process before this? Was it a candidate assessment, an internal talent review, something else? Name the specific pain: too slow, too rigid, inconsistent scoring, poor completion rates — whichever is true.

What I built

TODO: Describe the agent itself. What does a session actually look like? What does it ask, and how does it adapt? Who used it, and how often?

What happened

TODO: Outcome, ideally with a real number — completion rate, time saved, adoption, qualitative feedback from stakeholders. If it's still early, say so honestly rather than inflating it — "in pilot with X team" is a legitimate status.

What I’d do differently

TODO: One honest note on a trade-off or limitation. Case studies that only show wins read as marketing; one real caveat builds more credibility than it costs.