• by Laurie Sullivan , Staff Writer, June 26, 2026

When Hyundai went looking for high-intent auto shoppers, it found a solution to lower costs by working with AI models and something a little technical known as cloud-based, known as containerization, which take ad buying into the supply-side platform (SSP).

This type of technology can reduce the price of ad impressions and allow advertisers to gain better visibility. 

More companies like Amazon are starting to provide this type of transaction and containerization.

In early June, IAB Tech Lab announced a project to standardize container technology for Open RTB specs.

Hyundai participated in a programmatic containerization pilot between Chalice AI and OpenX of the SSP’s container product, called OpenXBuild, which launched this year. The process lets the advertiser see more of how the cost per programmatic impression.

This partnership between Chalice AI and OpenX introduced a new approach to video curation, where custom AI decisioning and bidding logic are done at the supply layer in real time, using OpenXBuild’s Real Time Bidstream API.

Canvas Worldwide supported the campaign and drove AI-driven decisioning into the supply path. 

The campaign drove more customer actions for less money and drastically reduced the cost of showing ads compared with any other type of strategy. The data shows 20% lower cost per high-value action and 67% lower CPMs compared with the next-best approach, according to data seen by MediaPost.

Adam Heimlich, Chalice AI CEO, told MediaPost these results are not “an isolated incident,” because advertisers are more often bypassing traditional platforms to leverage AI.

Heimlich came from agencies and witnessed firsthand the problem of "tooling for big advertisers" not being able to accomplish the tasks. It took Chalice AI's technologists and engineers five years to build the process and the technology that also supports customers like Bayer, and Zillow.

"It's really a movement to adopt this type of technology," Heimlich said. 

For Hyundai, the AI sits inside OpenX’s servers accessing the data direction by using an IAB standard called the “Agentic Realtime Framework,” Heimlich said.

The Chalice AI and OpenX framework can take one advertiser’s data to provide bidding instructions based on predictions and value to Hyundai, one advertiser at a time.

“It predicts the value of ad opportunities specific for one brand and its goals,” he said. 

Rather than layering AI on top of a buying transaction, this moves the automated process of evaluating, choosing, and pricing ad space in real time into the supply part of the process. It allows advertisers to optimize impressions in real time against actual business results, such as a completed purchase or a qualified lead.

Real-time bidding decisions and impression drove gains at the point where ad inventory sits--at publishers, websites, apps, and connected TV (CTV) networks.