As PM #3 on Qualtrics' Agentic AI incubation team, I lead strategy for agent testing, versioning, preview, and knowledge base capabilities for customer-facing AI agents at enterprise scale.

I replaced spreadsheet-based testing with a structured framework combining synthetic, representative, and regression testing. It used to eat up 41% of implementation hours and $71K per launch phase.

I also replaced ad-hoc agent cloning, more than 50 duplicates at its peak, with a proper versioning and rollback model, and aligned our preview, testing, and monitoring environments so nothing gets lost between them.

I led the 0→1 development and GA launch of Personalized Action Recommendations, EX's first manager-focused recommendation product. It's now used by more than 100 enterprises.

I also built the multi-language support and role-based access control underneath it, making this the first Qualtrics AI feature to launch with full support across all six tier-1 languages.

I expanded Conversational Survey AI from CX into EX in a single quarter. Built on response clarity and adaptive follow-up, it's now EX's fastest-growing AI feature.

Alongside that, I incubated EX's agentic AI roadmap from zero to one, working with five enterprise customers to narrow seven candidate use cases down to three prioritized bets.