Virtual Product Placement (VPP) is a computer vision-driven advertising product that enables brands to be contextually inserted into videos in post-production, creating new advertising opportunities without disrupting viewer experience. In this session, we’ll share a practical case study on how we took VPP from an initial proof of concept to a production-grade product. We’ll start with the business problem: why traditional ad formats can hurt the viewing experience and why post-production placement is harder than it looks given studio constraints, brand safety and content variability. We’ll then walk through how we defined which product aspects should use computer vision/ML versus rule-based approaches, set success metrics (realism, safety, scalability, throughput) and partnered with computer vision/ML scientists to meet those goals. We’ll cover human-in-the-loop design, cost/quality trade-offs and the operational steps required to scale across large content libraries. We conclude with the business impact and lessons learned.