Box introduces new AI agents for content management

Box introduces new AI agents for content management

Box launches Box Extract, Box Automate, and improved Box Apps to automate business processes with AI agents. These solutions are designed to help organizations extract data, orchestrate workflows, and build intelligent dashboards.

Box Automate enables organizations to orchestrate workflows between different agents and teams. The platform supports everything from routine task automation to fully autonomous, cross-system workflows.

The visual builder enables the design of workflows without requiring programming knowledge. Users can create custom agents for specific needs, such as Q&A, Compose, Extract, and Research agents. These agents can dynamically route tasks between people and systems based on business logic.

Integration with Box-native tools such as Box Forms, Box Doc Gen, and Box Sign enables end-to-end automation. Workflows can also be extended to custom applications and external systems via APIs.

Data extraction with AI agents

Box Extract is an important pillar of the new agentic solutions. The system utilizes advanced AI to extract data from various document types, including contracts and scanned forms. To achieve this, the company utilizes both Standard and Enhanced Extract Agents, which can comprehend documents and extract relevant information from them.

The tool can handle PDFs, spreadsheets, images, and handwritten text. By recognizing semantic relationships between fields, the system also extracts complex, interconnected data points. Built-in validation ensures that the output is immediately usable for business workflows.

Intelligent dashboards with agent assistance

Box Apps is getting new AI capabilities to manage business-critical processes better. The no-code dashboards now combine content, metadata, users, workflows, and agents in a single environment.

Users can use natural language queries to filter and explore content quickly. The system generates dynamic data visualizations such as graphs for actionable insights. Agent-assisted analysis helps identify trends, anomalies, and recommended actions.

Multiple dashboards can be customized for different teams. Embedding in third-party platforms such as Salesforce ensures seamless access and interaction.

Flexibility in AI models

Box maintains flexibility by giving customers a choice of different AI models. Through Box AI Studio and Box AI APIs, organizations can choose from state-of-the-art models from Anthropic, Amazon, Google, IBM, Meta, OpenAI, and xAI.

In addition, customers can extend Box content to third-party agents via the remote Box MCP Server. Existing integrations with Anthropic Claude, Microsoft 365 Copilot, Mistral Le Chat, OpenAI ChatGPT, and Salesforce Agentforce are available, with GitHub Copilot, IBM WatsonX Orchestrate, and ServiceNow AI Agent Fabric in development.

The three agentic solutions will be available to customers with Enterprise Advanced subscriptions in the coming months.

Tip: Box aims to take collaboration in hybrid world to next level