Case study 01 · 2022 — Present
Rebuilding a demand-side platform, AI-first.
The problem
Nexxen's DSP had grown through acquisition. Every part of the platform looked and behaved differently, campaign optimization was slow, and QA consumed hours that should've taken minutes. Customers noticed.
The approach
Built and trained the PM team, defined the DSP vision end-to-end, and ran a multi-year redesign on Vuetify components. Embedded agentic AI directly into campaign setup, QA, and troubleshooting — not as a bolt-on, but as a first-class part of the workflow.
Embedded agentic AI capabilities directly into user workflows to enhance QA checking and troubleshooting efficiency.
Flyout panels in action — edit, analytics, audit log, and bid exclusions in a single context-preserving surface.
What shipped
- New pacing and projection visualizations for in-table decision-making
- Flat View to compare tactic performance across advertisers
- AI-powered analytics and deal troubleshooting, embedded in context
- Redesign of whole layout and workflows
What I led
- Built the PM organization and defined vision, strategy, and roadmap
- Installed Design Thinking and Agile across three distributed teams
- Deployed Lovable.ai, Pendo, Jira, Coda, Confluence as the working stack
- Ran the program from concept → MVP → go-to-market
New UI product vision walkthrough — the full DSP redesign from first principles.



