TwelveLabs expands video AI technology on AWS Marketplace

Large-scale language models grab most of the headlines, but what about video AI?

TwelveLabs Specialties Multimodal AI model You can “watch” and analyze video content. Established in 2020the company has built a strong customer base around its video intelligence capabilities.

“Currently, over 80% of the world’s data is video, yet it is the least understood.” Danny Nikolopoulos (Photo), Head of Strategic Partnerships at TwelveLabs. “Naturally, the first ones to step up were the companies and industries that monetize video, and that’s primarily what we call TMEGS, which is Telco, Media, Entertainment, Gaming and Sports. We’re partnering with most of the big players in that industry right now.”

Mr. Nikolopoulos talks with theCUBE Christophe Bertrand During an exclusive interview on theCUBE, SiliconANGLE Media’s live streaming studio, for the AWS Marketplace series. They discussed what video AI brings to the entertainment and sports space and what TwelveLabs has achieved. expanded the range Via AWS Marketplace. (*Disclosure below.)

Video as a new frontier for AI

TwelveLabs offers two AI models. Marengo models are used to search videos in the same way humans watch videos and recall memories. Pegasus, on the other hand, is a video language model that allows you to analyze video content to gain further insights. Together, these models form what Nikolopoulos describes as a “powerful librarian.”

“The example I’d give is if you’re with a friend and you say, ‘Hey, remember that play in the last game where the quarterback threw a touchdown?'” he said. “If you immediately play it back in your mind, you can see it happening. This is how our model works. It allows us to really pinpoint moments deep in the video that contain semantic language that you wouldn’t be able to search for any other way.”

In addition to Marengo and Pegasus, TwelveLabs recently released Rodeo, an AI co-pilot that helps video editors find and assemble footage with natural language prompts. This set of features has proven useful for customers such as a Canadian sports company that uses TwelveLabs’ model to create short highlight reels for social media.

“Rodeo was born in response to customer demand,” Nikolopoulos explained. “[They] They were considering using models and building models independent of each other, but specifically wanted it for more specific purposes. It was, “How can we actually accelerate the editing process to not only speed up search and analysis, but also give us a head start on what is a very difficult process?”

By joining AWS Marketplace, TwelveLabs has reached a broader customer base and discovered additional uses for its model beyond the typical entertainment and media sectors. Customers often already have data they want to analyze on AWS, but even if they don’t, TwelveLabs has a number of integration partners that are part of the AWS ecosystem.

“We’re finding more use cases,” Nikolopoulos said. “For example, a major automaker has built an automated mechanic that inspects cars that come into the store and performs an initial analysis. Law firms are trying to make sense of large volumes of discovery and investigative evidence. Insurance companies are trying to replace paper-based processes with video-based processes that require less human intervention.”

Below is the full video interview, part of SiliconANGLE and theCUBE’s coverage of the AWS Marketplace series.

(*Disclosure: TheCUBE is a paid media partner for the AWS Marketplace series. Neither AWS, the sponsor of theCUBE’s event coverage, nor any other sponsor has editorial control over theCUBE or SiliconANGLE content.)

Photo: SiliconANGLE