AI-Driven Product Placement at the Speed of Live

Overview

THE CHALLENGE

Tagboard needed a way to help producers act on inspiration in the moment—without breaking the flow of live production—as the current process of identifying the right product during broadcasts was labor-intensive, error-prone, and costly, requiring either remarkable foresight or manual support.

THE SOLUTION

Tagboard developed a powerful feature within its interactive production app that enables affiliate marketing during live TV and streaming broadcasts, using a dynamic QR code template to seamlessly direct viewers to purchase pages for featured products at key moments.

THE OUTCOME

Ollion’s AI engine enabled real-time product recommendations, reducing manual effort and boosting engagement, conversions, and cost-efficiency—leading Tagboard to expand Spark AI across its platform.

About the Business

Tagboard is a leading cloud graphic system powering interactive storytelling for the world’s biggest broadcasters, sports teams, and content creators.

Their innovative tools enable producers to build dynamic, interactive segments and streamline production workflows for live and on-demand content. Used by more than 500 media brands globally, Tagboard is on the frontlines of real-time audience engagement and broadcast innovation.

The Business Challenge

Tagboard developed a powerful feature within its interactive graphics system that enables affiliate marketing during live TV and streaming broadcasts. Using a dynamic QR code template, producers could direct viewers to purchase pages for featured products during key moments of a broadcast. However, to capitalize on those moments, producers needed either to pre-select products in advance or identify the right product in real time.

Doing so live required either remarkable foresight or a team of interns conducting manual searches—a labor-intensive, error-prone, and costly workaround. Tagboard needed a way to help producers act on inspiration in the moment, without breaking the flow of live production.

The Ollion Solution

To solve this, Tagboard partnered with Ollion to design and implement Spark AI—an embedded, real-time recommendation engine that automates product discovery during live broadcasts. The AI-powered engine integrates directly with Tagboard’s app, engaging with producers to build intelligent user and team profiles. By combining these profiles with contextual broadcast data, the solution taps into supported product catalogs, conducts relevance-based searches, and delivers optimized product suggestions—all in real time.

At the core of the solution is Amazon Bedrock, using agents from providers like Anthropic to interpret broadcast cues and return targeted results. These agents are orchestrated through AWS Lambda functions, which manage workflow execution and serve responses to the application interface. AWS S3 handles data storage, session management, and contextual memory.

The entire solution is serverless, enabling high-speed interactions with a cloud spend of less than $1 per multiple full workflows executed. In seconds, Spark AI evaluates thousands of candidates, empowering producers to instantly surface relevant products and dynamically trigger affiliate links for viewers.

Go behind the scenes of Spark AI and its impact on live media — read more from Tagboard: Introducing Tagboard’s Spark AI & Spark AI: Your Agentic Production Assistant

With Tagboard's Spark AI for Shoppable, we’ve reimagined how live production teams interact with content and commerce in real time. It’s more than just automation—it’s about empowering storytellers with intelligent, intuitive tools that help them move at the speed of live. Partnering with Ollion allowed us to bring this vision to life quickly and seamlessly, with a scalable solution that truly understands the broadcast environment.

Christine Chalk, Chief Product Officer, Tagboard

The Business Benefits

With Ollion’s AI-powered recommendation engine in place, producers can now surface targeted products at exactly the right moment—without needing to pre-plan or rely on a swarm of live researchers. This shift reduced the manual burden on production teams, increased agility, and improved both engagement and conversion rates tied to affiliate marketing efforts.

The new system executes thousands of product discovery requests in real-time—often under $1 per broadcast—with measurable improvements to performance, accuracy, and cost-efficiency. Tagboard now plans to expand the Spark AI capability across their full product suite, using both qualitative and quantitative user data to continuously refine the experience.