sevDesk

I built a semi-automated customer feedback evaluation process for sevDesk, turning the daily stream of support tickets, app store reviews, and complaints into structured input for product and design. The system reads incoming feedback, clusters it by theme, and breaks the noise down into user stories, feature requests, and initiative-level patterns — with a human in the loop for the judgment calls AI shouldn't make alone.

Product managers stopped spending their Mondays triaging spreadsheets and started them with a ranked view of what customers actually want. Feedback that used to die in a queue now travels directly into the roadmap with its context intact.