Egnyte adds generative AI to streamline content collaboration

Today, content collaboration platform Egnyte announced new generative AI capabilities to simplify data access for its customers.

The California-based company said it is using private instances of various AI models to provide users with summarization, transcription and search capabilities across multiple file types. The move will enable enterprises to find and synthesize information buried in a massive trove of documents and media files much more quickly than ever before.

However, as of now, these features are rolling out to only a limited set of Egnyte users, with broader access likely at a later stage.

Content access simplified with generative AI

Since its launch over a decade ago, Egnyte has evolved into an enterprise-focused file-sharing solution that allows remote teams of all sizes to share and collaborate on their confidential data assets — while keeping privacy and security at the forefront. The company claims to have democratized cloud content security for more than 17,000 organizations across sectors.

Now, as the next step in its evolution, Egnyte is adding generative AI smarts into the solution to make sure every enterprise executive using the platform can quickly get the information they need. At the heart of this effort are three new developments: the ability to summarize the information contained in complex documents; the creation of text-based transcripts of audio and video files; and the ability to find photos in the library that contain a particular object.

As David Spitz, chief strategy officer at Egnyte, told VentureBeat, these features save users from the hassle of moving their files out to a different software system just to extract value from them. This could have the benefit of preventing violation of corporate policies while keeping the data in question safe at the same time.

>>Follow VentureBeat’s ongoing generative AI coverage<<

Along with the summarization and transcription capabilities, Egnyte will also provide users with an AI chat interface, a way to ask quick questions about any document opened on the platform. 

“For example, when opening up a contract, a user can ask for terms of payment, termination clauses, or services provided. Users can gain time instead of sourcing through the agreement to grab those details,” Spitz explained.

What goes on the backend?

Under the hood, Egnyte is using the GPT-3.5 and GPT-4 large language models from the Azure OpenAI service. The company is using the private instances of these models, along with its content governance framework, to ensure that both the source data and the AI-generated responses adhere to the security and compliance policies of customers using the platform.

“We have added proprietary mechanisms to verify that the results are accurate and appropriate. We also worked closely with Microsoft to ensure that no customer data could be shared with the foundational models or other customers, ensuring that any information generated by the models would respect the access controls of the customer,” Herb Roitblat, lead data scientist at Egnyte, told VentureBeat. 

Notably, the features also cite the source of the information provided, making users contextually aware of where the information is coming from. 

More to come

While Egnyte has been making use of AI for quite some time, helping with things like data protection and detection of anomalous usage patterns, the latest move makes sure that the technology is accessible to every user of the platform. 

As a next step, the company plans to build out these efforts by making the features announced today generally available and launching new AI capabilities to simplify how teams collaborate on their content.

“After this first phase of AI enhancements within Egnyte, we plan to continue to expand on generating synthesis and insights to help customers glean more insightful responses to their queries. Egnyte will continue to make it easier for people to work,” Spitz said.

>>Don’t miss our special issue: The Future of the data center: Handling greater and greater demands.<<