Box’s AI agents advance your business’ A2A vision

The development of agent-to-agent protocols such as A2A has the potential to fundamentally change the way businesses operate.

Box Inc.’s current artificial intelligence strategy is built on interoperability. Intelligent content management platform partners with Google Cloud Create your own agentfocuses on data extraction and classification, and workflow automation.

“Ninety percent of the content that organizations own, whether it’s HR documents, product documents, or financial documents, is unstructured. That means the heart of a company that runs on content is difficult to access and use.” Yashoda Bhavnani (Photo), Head of AI at Box. “AI basically just unlocks this.”

Bhavnani spoke to theCUBE John Farrier for Google Cloud AI agent hands-on series on theCUBE, SiliconANGLE Media’s live streaming studio. They discussed A2A and the collaboration between Box and Google Cloud. (*Disclosure below.)

Transform unstructured data into insights

Box leverages Google’s Gemini model and Vertex AI platform to develop agents such as Box Extract to perform deep data extraction and Box Automate to manage business workflows. Bhavnani said one of the company’s biggest goals is to make unstructured data feel structured.

“If you have a 100-page rental agreement and you have multiple copies of it, and you want to drive to the house and say, ‘When was this written? When does it end? What is the value of the agreement?’ Now, instead of parsing 100 pages, AI can extract that information,” she said.

Gemini’s multimodal capabilities help you achieve this. One use case that Bhavnani highlights is insurance claims files. This is proprietary and contains a lot of unstructured data. Box Extract can analyze thousands of those claim documents and identify specific information such as end date and whether the claim was accepted.

“They no longer have to sit there and go through 1,000 documents and create spreadsheets,” Bhavnani said. “These insights are securely attached to documents in Box, so they can be easily queried. You can ask, ‘What were all the claims submitted in the last month in the X thousand price range?'” And that’s gone from hours to minutes in some cases. ”

The future of A2A

Box sees itself as a “valuable link in the chain.” The future will not be a single agent working in a silo, but an agent working across an entire ecosystem. Bhavnani believes that using the A2A protocol with Gemini Enterprise is important because it allows enterprises to use Box AI with other agents in a cross-ecosystem.

“The best agents have the best context, and the box has it,” Bhavnani said. “Box has knowledge, it has all the knowledge across the enterprise. If you think about it that way, you can use AI, you can build agents, you can use agents in this context. The richer that context is, the better the agents are.”

A2A is the next frontier in business workflows, with dedicated AI agent department Regarding Google Cloud Marketplace, Bhavnani added: These platforms streamline reseller partnerships and agenttic AI product sourcing, allowing companies to adapt agenttic AI to their workflows with their unique architectures and costs in mind.

“When we bring the Box AI agent to the Gemini Enterprise AI Agent Marketplace, what we are really offering is enabling our customers to really harness the power of AI in their content while having direct access to their own Google Cloud environment,” said Bhavnani.

The marketplace also allows businesses to safely interact with Box AI. A2A only works if businesses trust their agents to work with sensitive data in a secure manner. Even Bhavnani admits that leveraging AI isn’t all sunshine, but he still believes the benefits far outweigh the costs.

“You have to have patience with AI,” she says. “I’ve seen the magic of building an agent, which is great. It works the first time. But I’ve also seen that there’s usually a sense of having to do it repeatedly, which means you’ll build the agent and say, ‘Oh, you didn’t provide me with the exact field that I wanted to repeat with 90% confidence. How can I re-prompt?’ So there’s a sense of patience, but it’s worth it. ”

Below is the full video interview. Part of coverage from SiliconANGLE and theCUBE. Google Cloud AI agent hands-on series:

(*Disclosure: TheCUBE is a paid media partner for the Google Cloud AI Agents in Action series. Neither Google Cloud, the sponsor of theCUBE’s event coverage, nor any other sponsors have editorial control over theCUBE or SiliconANGLE content.)

Photo: SiliconANGLE