There were 687 press releases posted in the last 24 hours and 391,963 in the last 365 days.

GrayMeta, Inc. launches Curio Anywhere, artificial intelligence for media files that runs on prem or in the cloud

GrayMeta introduces Curio Anywhere, a machine learning data platform for your enterprise.

Tell more compelling stories, generate new streams of revenue, monetize your archive, and more with Curio Anywhere running on prem or in the cloud.

This is going to be game-changing for any customer with a lot of media content. You can deploy suite of machine learning services anywhere, tune with your own data, and integrate into any application.”
— Aaron Edell, CEO of GrayMeta
BEVERLY HILLS, CA, UNITED STATES, April 13, 2023/ -- GrayMeta, Inc. today announced Curio Anywhere, bringing customers the ability to automatically index all their media assets using machine learning entirely on premises or in their own cloud environment. Once Curio Anywhere is deployed, customers can add the storage locations for all their media content, and Curio will begin processing each one with an array of tunable machine learning models such as face recognition, logo recognition, image tagging, speech to text, language detection, optical character recognition (OCR), technical cue detection, and more. GrayMeta has handcrafted state of the art machine learning models designed to run on premises, or locally to the Curio environment, so no content is getting posted to a public cloud endpoint. The metadata generated is then stored in a searchable database, accessible through Curio's elegant UI or world-class API. Some of the machine learning models can be further trained, or tuned, within Curio’s AI Studio experience using their own content. These finely tuned machine learning models will increase accuracy and help tackle new use cases.

Broadcast media and sports organizations spent $30 billion in 2021 creating media content. Considering the high cost of producing and acquiring the rights to media, content owners need to find ways to monetize what they own. For many customers, these archives can grow to tens of millions of hours of content stored in various locations. Customers are unable to know what content they have, where it exists, and how to further monetize it because there’s no single source of truth that contains all the metadata about those files. Some customers who have migrated to the cloud are having to download entire directories just to see what files are in them, incurring expensive egress fees in the process. Other customers are having to reshoot content because they can’t find the originals. Customers need more than just search, they need to analyze their content so they can power better recommendation engines, enable improved ad placement, report on brand awareness and sponsorship effectiveness, and more. Machine learning allows these customers to automate what would otherwise be a laborious task of tagging every second of every video with metadata. However, most solutions on the market today rely on sending content out to public cloud endpoints. But this has been cost prohibitive for most customers, in some cases costing as much as $15 per hour of content processed. Furthermore, content owners may have security or legal restrictions on sending content outside of their data center or country. These cloud-based models are also trained on generic sets of data, and can have accuracy issues for some customers. Customers might also have custom needs for detecting visual categories, objects, phrases, or locally known people, requiring a way to train or tune the machine learning models inexpensively and quickly.

GrayMeta’s new Curio Anywhere platform solves these challenges and more by providing a solution that automates applying machine learning models to media files all within a customer’s environment. Whether Curio Anywhere is running on a single appliance, a local cluster in a data center, or in a customer’s cloud environment, the machine learning models don’t need to make public API calls to anywhere - everything happens locally. With Curio’s AI studio, customers can teach, tune, or train the models on their own content, improving accuracy and enabling new detection use cases.

Although Curio Anywhere has a robust user interface with search capabilities, frame-accurate playback, data export features, and more, it also has a world-class API, and upon launch, will be available with a connector to your favorite business intelligence (BI) tools. Customers can use pre-built or customize their own Amazon QuickSight dashboards, for example, to quickly locate subsets of content that are in the wrong aspect ratio, filter by languages spoken, quickly see a list of content that is flagged for inappropriate language or nudity, generate reports on the amount of time logos or brands appeared on air and correlate it to analytics about revenue, and their recommendation engines more data to be trained on for better personalization and higher engagement. With Curio Anywhere, anyone can spin up a suite of machine learning capabilities that automatically processes files and populates a database anywhere they want.

Dan Daube
Visit us on social media:

Brief overview of Curio