Box, Inc., the leading Intelligent Content Management (ICM) platform, has announced the general availability of Box Extract, a new capability designed to help enterprises unlock value from unstructured content at scale. Powered by leading generative AI models from Google, Anthropic, and OpenAI-and enhanced with advanced agentic capabilities—Box Extract enables organizations to securely extract critical information from content and store it as structured metadata within Box.
By transforming unstructured documents into usable, actionable data, Box Extract makes it easier for enterprises to automate workflows, accelerate decision-making, and surface insights faster across their organizations. The solution allows content to actively support business processes rather than remaining locked in static files.
“Enterprises are sitting on a gold mine of data in their untapped content,” said Aaron Levie, co-founder and CEO of Box. “With Box Extract, that information is now unlocked and can transform how businesses analyze information and make decisions. By turning unstructured content into structured, usable data, organizations can deliver real-world impact by having their content actively work for them across their most important lines of business.”
“In the financial services industry, we work with highly sensitive data that demands the highest levels of security,” said Geoff Moore, CIO at Valmark Financial Group. “Box Extract pairs enterprise-grade security controls with the deep subject matter expertise of our employees to extract data from unstructured sources and transform it into actionable insights securely. By mining information from sources such as account forms, insurance illustrations, and commission statements, we’ve achieved exceptional gains in both efficiency and accuracy.”
Also Read: PhotoShelter Unveils Lumen Portal UI to Modernize How Organizations Share Visual Content
“Every day, we process vast amounts of unstructured documents that are critical to serving Texans efficiently,” said Wendy Barron, CIO at Texas Department of Motor Vehicles. “With Box AI, we’re now able to automatically extract key information from forms and records, reduce manual review, and accelerate our workflows, all while maintaining the security and compliance standards required of a public agency. Implementing modern AI tools has helped our team focus less on paperwork and more on delivering timely services to the public.”
Box Extract
Across organizations, valuable institutional knowledge is hidden in a sea of contracts, product specifications, policy documents, reports, charts, and other forms of unstructured content generated in day-to-day operations. This information gives context to all AI models and agents for the delivery of meaningful business outcomes. Traditionally, the extraction of value from unstructured data has depended on both manual review and legacy technologies that are expensive, burdensome to maintain, and hard to scale.
Box Extract overcomes these limitations by combining advanced generative AI models from Google, Anthropic, and OpenAI with sophisticated agentic intelligence. These come together to provide the power of accurate extraction from complex, multi-format documents with enterprise-grade security and governance.
Unlike classic tools that just extract crude text, Box Extract uses an agentic approach to understand document structure and semantic meaning. Box Extract is able to break the content down into its constituent parts, such as paragraphs, tables, and charts, before then identifying and extracting just the most relevant information. It’s also possible for an organization to create custom Box Extract Agents to meet their particular business needs and to deploy them securely across large content repositories.
Extracted data is stored alongside content in Box as custom metadata, and can be exported or synchronized with downstream systems including Databricks and Snowflake. This allows enterprises to seamlessly integrate structured insights into broader analytics, automation, and data workflows.
With Box Extract, organizations can:
- Make faster, more informed decisions using metadata-powered dashboards and views in Box Apps
- Automate end-to-end workflows using Box Relay today and Box Automate in the future
- Improve content discovery and accelerate search experiences for all Box users
- Extend metadata into third-party and custom applications
Applying Box Extract Across Industries
Box Extract supports a wide range of industry use cases, including:
- Financial services: Loan creation and servicing:Extract due dates, terms, and payment information in financial services for accelerated reconciliation and approval
- Government & Public Sector: Unify compliance procedures and public service delivery through the extraction of types, costs, dates for inspections, and other important elements from public records
- Media & Entertainment: Extract automatically titles, authors, rights owners, versions, and scene keywords from scripts, contracts, and creative content to optimize content discovery and rights management
- Insurance: Increase accuracy and reduce manual processing of claim reviews and determinations by extracting key information from accident reports and medical bills
- Legal: Enhance contract management by identifying clauses, counterparty information, renewal terms, expiration dates, and obligations within long-form agreements
“In today’s market, the speed of your data drives business results. By integrating Google’s Gemini into Box Extract we’re helping customers instantly turn massive amounts of content into structured, actionable intelligence,” said Matt Renner, President and Chief Revenue Officer at Google Cloud. “Gemini’s power to process and understand complex, high-volume data allows customers to move beyond simple scanning to true workflow automation, dramatically speeding up critical processes like loan approvals and contract management.”