Cineverse leveraged Google Cloud to create an AI-powered search and discovery tool, which includes its own AI “video advisor” called Ava, that’s supposed to help ease the consumer pain point of finding TV and films across streaming services.
Named cineSearch, the platform is debuting in beta this spring, initially available within the company’s namesake Cineverse streaming platform on the web. Users can sign up for the public beta here, starting today. In the near future the company plans to make it more widely available by licensing it to OEM and third-party streaming platform partners. As part of the platform Cineverse is introducing what it calls a new “artificial intelligence-based video advisor” called “Ava.” While “Ava” will be rolling out on Cineverse, the cineSearch platform is white label, meaning partners that incorporate the technology don’t need to retain the Ava name or branding.
In announcing the platform, Cineverse said it aims to answer viewers’ key question of: “What do you want to watch?”
In its initial phase the service is only supported through desktop text-based queries, but Cineverse plans to make it voice-enabled when releasing the platform on connected TV, a Cineverse spokesperson confirmed to StreamTV Insider. That means that at first, it will only be supported on the web-based version of the Cineverse streaming service, expanding to mobile afterwards.
Cineverse said the tool, which promises a conversational experience for those looking for advice on what TV shows and movies to watch, was created by the company using Google Cloud’s Vertex AI Search infrastructure.
“The underlying core cineSearch technology was built as an adaptive layer on top of the PaLM 2 Large Language Model (LLM) with the intent to craft a chatbot specifically tuned for feature film and television entertainment,” the company described in a press release.
Also underpinning the technology to tailor the what the company called “the first proprietary movie industry-specific AI model” is an extensive set of metadata, which it described as a combination of standard metadata alongside computer vision-based enriched, contextual metadata like weather or setting – with the model spanning more than 100,000 Hollywood movies and TV shows. Cineverse declined to share its metadata partners but said it has various deals in place with key traditional providers, AI-based metadata providers as well as open source.
CineSearch supports dozens of search dimensions, including theme, tone, mood, setting, music score, plot, and micro-genre, among others. The platform is also pulling other signals from users to determine the best titles to surface. Cineverse said the service can leverage scene-specific metadata and a viewers’ watch history, alongside their location, the current date, local weather conditions and other inputs “to determine the ideal titles to recommend.”
Here are two examples of queries cineSearch can handle:
“Can you recommend a quirky rom-com with a good soundtrack?”
“Can you recommend a suspenseful drama with a complex characters and a plot twist?”
And importantly, cineSearch isn’t just supposed to improve finding content on the Cineverse or app of choice but is meant to provide a unified search engine across streaming services. If a user’s query matches titles that aren’t available in the Cineverse library, it will recommend relevant titles from other streaming services and provide direct links to access them.
“Effective search and discovery is currently the most pressing problem for users of streaming services today,” said Tony Huidor, CTO and COO of Cineverse, in a statement. “We first developed cineSearch as an answer to our own problem. Using Google Cloud’s Vertex AI platform, we now have the ability to expand this feature well beyond our initial expectations. We feel this is a great example of leveraging the power of AI to not only create better user experiences, but to also help aid in the discovery of great films that fans may not have otherwise found.”
Anil Jain, global managing director of Strategic Consumer Industries at Google Cloud, in a statement said that “generative AI has the potential to fundamentally transform the media industry,” with cloud capabilities that can enhance the consumer experience.
“By leveraging Google Cloud’s leading gen AI capabilities, cineSearch can further personalize and streamline the user experience, providing viewers a vastly improved way to discover new content, with relevant recommendations for movies and television shows at their fingertips,” Jain continued.