OtherLevels rewrites the game with hyper-personalized fan experiences

Under the surface of these seamless and immersive fan experiences lies a complex sequence of data-driven components brought together by a generative AI layer. OtherLevels matches the context of a game with behavioral information to create a rich, personalized view of the customer before generating content that resonates with their interests. “We’re really able to drill down into what makes that person tick,” says Jordan Scott, Software Engineer at OtherLevels.

This highly personalized and localized approach makes it possible to cut through the content noise surrounding fans. “They aren’t just getting blanket coverage, they’re getting meaningful content at the right moments,” says Graham Merrett, CTO at OtherLevels. Whether it’s a hole-by-hole golf replay or detailed statistics on their favorite football players’ performance, brands can pinpoint when to engage fans and provide added value—all in real time before the opportunity passes.

For example, in a mobile game, users receive a push notification before a character video pops up. The brand’s digital ambassador then shares a timely, relevant message with a game highlight that encapsulates the buzz of the moment. As Merrett explains, “It might say, ‘Hey Graham, your team were behind and now they’re making a comeback—watch this goal!’” Fans can then get to the heart of the action by jumping back into the app.

As a small company with big aspirations, OtherLevels brought its vision to life in less than three months by rapidly testing and refining ideas with the Amazon Prototyping and Cloud Engineering (PACE) team. “Partnering with Amazon Web Services (AWS) means you just move at a different velocity,” comments O’Kane. “The level of attention, support, and access to expertise is second to none.”

The AWS Builder Studio also gave the company the space to experiment with cutting-edge generative AI tooling and deep dive into its strategy and architecture. “Our original digital twins were fun, but they just weren't engaging enough,” says Scott. Working hand-in-hand with AWS to ideate and articulate their solution also meant that the company could significantly cut time to value and minimize technical risk without heavy investment. As O’Kane notes, “The AWS Builder Studio amplifies solutions. We can deliver outcomes faster and get them in front of brands.”

Giving life to digital twins with a human look and feel involved refining speech, movements, and lip synchronization. “Amazon Bedrock foundation models (FMs) have allowed us to really utilize the best generative AI models available,” Scott adds. To create a diverse set of characters, OtherLevels trains these FMs with Amazon SageMaker using domain-specific inputs. Once it has generated the videos, high-quality human voices are synthesized through a text-to-speech component using Amazon Polly. The result is convincing characters that can interact using fluent speech and fluid movements. “Now we can customize an ambassador that’s 100 percent on brand,” says Scott.

By running on AWS, OtherLevels can rapidly spin up resources as they face spikes in traffic, including during hotly anticipated games. It also has the foundations to explore new products and industry applications, such as in the travel sector. As Merrett explains, “The opportunities go way beyond the space we’re working in. It can be applied to any customer interaction to reach them quickly, in a personalized way, and at scale.”

As generative AI continues to develop at breakneck speed, the OtherLevels team is taking advantage of the tools and support available to help them innovate and scale on their terms. “Working with AWS has really opened up this big playground and ability to approach a problem from a different angle,” says Scott. With the freedom to think bigger, their technology is making them a frontrunner in one-to-one customer experiences.