Wasabi Acquires GrayMeta’s Curio AI To Take Unstructured Data Out Of The ‘Dark Ages'

‘Imagine walking into Widener Library at Harvard with 11 million volumes, and there's no card catalog. That's what we have right now with unstructured data in the cloud. And I think our acquisition of this machine learning technology, along with [GrayMeta President and CEO] Aaron Edell and his crackerjack AI team joining Wasabi, is really going to be the most important development since the introduction of object storage itself,’ says Wasabi CEO David Friend.

[Editor’s note: The original version of this story inaccurately described the nature of Wasabi’s acquisition. Wasabi is only acquiring technology developed by GrayMeta.]

Cloud storage developer Wasabi Technologies Tuesday unveiled its acquisition of Curio AI, a technology developed by California upstart GrayMeta that uses AI and machine learning to automatically generate a searchable index of unstructured data.

The acquisition of Curio AI from Beverly Hills, Calif.-based GrayMeta, for which no dollar value was given, brings Wasabi the technology that will allow its users to easily search their huge archives of unstructured data, something that was not possible before, said David Friend, CEO of Boston-based Wasabi.

“Imagine walking into Widener Library at Harvard with 11 million volumes, and there's no card catalog,” Friend told CRN. “That's what we have right now with unstructured data in the cloud. And I think our acquisition of this machine learning technology, along with [GrayMeta President and CEO] Aaron Edell and his crackerjack AI team joining Wasabi, is really going to be the most important development since the introduction of object storage itself.”

Wasabi has customers storing multiple thousands of hours of video in Wasabi and can't find anything, Friend said.

“We've pretty well beaten this problem in strusctured data with SQL databases and things like that,” he said. “But right now unstructured data is still in the dark ages. And I believe that what we're doing here with Curio AI to automatically create an index of every face, every logo, every object, every sound, every word, will really revolutionize the utility of object storage for the storage of unstructured data.”

GrayMeta was founded in 2014 as a developer of technology to abstract metadata from media files simply by scraping them, not with machine learning, but instead using the embedded metadata such as what camera was used to generate a video or the GPS coordinates of where a video was made, Edell told CRN.

“It was around 2016 when I first saw a demo of a machine learning service that would transcribe audio,” he said. “The accuracy was very poor, and it was expensive and very slow. But that's when the light bulb went off in my head where I thought, 'OK, we're not going to get a lot of metadata by trying to look into the embedded metadata of files. Where we're really going to see value is machine learning, because it's going to be able to watch or listen and give us metadata. And that transcript, even though it wasn't something you'd want to turn into closed captions or anything like that, surfaced keywords that were searchable at that point. That's the card catalog, right? So you have somewhere where you can start actually searching for and finding instances within your video content. And that was the problem we were trying to solve.”

The biggest unstructured files and data in the world are video files, and Curio was built around the idea of harvesting metadata from media objects including video and audio, said Edell. Curio was also built to work with any storage media customers were using to store their unstructured data, he said.

“We used to support all object storage or file system storage locations,” he said. “We were not specific to anyone until now.”

Friend said that Wasabi will fully integrate GrayMeta’s Curio technology into his company’s cloud storage offering, and not offer it as a standalone technology for other storage clouds.

“It's going to be one integrated product, and it's going to be sold by the terabyte just like our regular storage, but at a slightly higher price. And for that, you will get unlimited use of the AI. So if you want to rescan looking for something that you didn't look for the first time, you can do that. You could sit there all day long doing searches and whatnot. And that's all built into the price. So we're trying to make it super simple, just like we did with our regular storage, with no egress fees, no usage fees, none of that stuff.”

Curio will automatically scan anything that's put into Wasabi’s intelligent tier of storage and produce an index which can then be accessed using the Curio user interface and one of several popular media asset management systems including Iconik, Strawberry, and Avid, Friend said.

“Our storage would sit below those things,” he said. “And would look provide the whole index so that anybody who's doing editing or video production work or any of that sort of stuff will be able to find what they're looking for in their video in their video content. We expect to do just what we've done with our regular storage, meaning we'll go to market with channel partners who sell into the media and entertainment industry.”

As an example, Friend cited as a potential customer the Boston Red Sox, with which Wasabi has had a long relationship.

“So if you are like the Boston Red Sox, you would take all your 50 years of archival video, move it into this Wasabi cloud which will be our ‘Curio cloud,’ and it'll automatically be indexed,” he said. “If you want to find that one time when so and so hit a home run in a World Series or something like that, if you've got any kind of keywords that you can use or anything that you could tie it to, you go right to it because Curio creates a second by second index of 50 years worth of video.”

Friend said the term “Curio Cloud” is not yet the formal name of the offering, and that marketing details for the new technology have yet to be finalized. He said the product is already available as proofs of concept for some customers, and that formal release is slated for late first quarter or early second quarter of 2024.

Edell said that GrayMeta already had paying customers, and that his team will continue to support existing customers as best it can.

“But there's going to be a real benefit to moving over to Wasabi for them,” he said. “They will have a single vendor to deal with. And there's going to be probably some economic benefit as well, since the way that we're going to price it with Wasabi is going to be very different.”

The benefit to Wasabi is also significant, Friend said.

“We think it's a major advance in object storage. And five or 10 years from now, everybody will have this. But in our strategy of being hyper-focused on storage, I think we're moving into the second phase of Wasabi which is not just to have parity with Amazon, Google, and Microsoft in terms of functionality, but to now in one leap become the leader in terms of functionality. The hyperscalers can’t do what we’re doing with Curio. I mean, they have a toolkit, and you can assemble something like this if you have the time and money. But there's nothing equivalent to this that anybody else is offering as far as I know.”