neoAI Chat Expands Integration with Box

neoAI Inc. (Headquarters: Tokyo, Representative Director and CEO: Shunsuke Chiba, hereinafter "neoAI") is pleased to announce that its enterprise generative AI platform "neoAI Chat" has expanded its integration with Box, now supporting two methods: app connection via Box MCP server and external data linkage.

This enables both real-time search of Box documents according to user permissions and semi-synchronization of Box files into neoAI Chat for high-precision RAG (Retrieval-Augmented Generation), allowing flexible utilization of enterprise information assets accumulated within the organization according to the use case.

Background

As generative AI adoption progresses in enterprises, it is important for AI to naturally connect with existing business systems and document management platforms to become embedded in daily operations. In particular, many of the information sources referenced in daily work—such as internal regulations, business manuals, application procedures, and customer response materials—are accumulated as documents, and there is a growing need for environments where these can be searched and utilized safely and efficiently.

Box is used by many companies as a platform for securely managing, sharing, and utilizing important corporate content. Meanwhile, in document utilization through generative AI, there are both needs: to reference the latest information as needed, and to perform high-accuracy search and response across large volumes of documents depending on the use case.

To address these needs, neoAI Chat has expanded its Box integration methods and started offering two selectable connection approaches.

Overview of Box Integration

neoAI Chat offers the following two methods for Box integration:

1. App Connection via Box MCP Server

This method uses Box's search tool to search Box documents in real time according to user permissions. It is suitable for use cases where users want to explore necessary information on the spot from the chat UI, referencing the latest documents or permission-based information.

For example, for inquiries like "Tell me about internal regulations" or "I want to submit a transfer request, how should I do it?", it is possible to directly search related documents in Box and answer based on the content.

This method features searching Box in its latest state, realizing search according to user permissions and access to always-up-to-date information. It is suitable for tasks requiring freshness, such as on-site immediate search, reference to the latest documents, and ad-hoc inquiry responses.

2. External Data Linkage

External data linkage semi-synchronizes Box files into neoAI Chat and utilizes neoAI's high-precision RAG function for search and response. It is suitable for cross-sectional information utilization across large volumes of documents or tasks that emphasize search and response accuracy.

For example, based on document groups accumulated within the company such as employment rules and personal information handling regulations, it is possible to obtain more practical answers.

This method features semi-synchronizing Box files for RAG utilization, enabling the use of neoAI's high-precision RAG function and cross-sectional utilization of large document volumes. It is suitable for tasks requiring deep information utilization, such as internal knowledge search, answer generation emphasizing accuracy, and business support using document groups.

Value Proposition

With this update, neoAI Chat can now approach Box content from both aspects: access to the latest information and high-precision utilization.

- When users want to check the latest materials on the spot according to permissions: App connection via Box MCP server
- When users want high-accuracy search and response based on accumulated documents: External data linkage

Companies can select the optimal connection method according to their business requirements and operational policies.

This enables more effective utilization of information assets stored in Box across a wide range of business scenarios, including inquiry responses, internal knowledge search, regulation and manual reference, and application procedure guidance.

Comment from Noriyuki Sato, President and Executive Officer, Box Japan

"We welcome the strengthening of Box integration in neoAI Chat. Box is used by many customers as a platform for securely managing and utilizing important corporate content. With neoAI Chat now supporting both search via Box MCP server app connection and high-precision RAG utilization via external data linkage, we believe the information assets stored in Box can be utilized in a more business-oriented manner.

We look forward to further expanding content utilization that contributes to customer productivity improvement and business transformation through collaboration with neoAI."

Comment from Koshiro Terasawa, Director and COO, neoAI

"With this expansion of Box integration in neoAI Chat, we now support two methods: app connection via Box MCP server and external data linkage.

By combining Box, which is deeply embedded in daily operations, with neoAI Chat's agent functions, we can provide an AI experience that not only helps find necessary information but also supports actual task execution.

neoAI will continue to support customers' business transformation through a generative AI platform that enables safe and practical utilization of enterprise information assets."

Future Outlook

We believe the key to maximizing the value of generative AI lies in how seamlessly it can connect with enterprise data. No matter how high-performing the AI is, it will not be truly useful if it cannot access the information assets necessary for work.

Therefore, neoAI Chat will continue to expand integrations with various systems actually used in enterprises, such as document management platforms, groupware, business systems, and knowledge bases.

Overview of neoAI Chat

neoAI Chat is an enterprise AI agent platform optimized for Japanese business operations, balancing AI governance through various management functions with the creation and operation of AI assistants by on-site users.

Key Features:
- Security for peace of mind: Multi-layered governance functions including SSO, IP address restriction, usage logs and permission management
- AI governance through diverse management functions: Supports administrator settings for assistants, prompt/model management, and usage log visualization
- UI/UX facilitating citizen development: No-code/low-code design enabling on-site departments to design and deploy AI assistants and agents themselves
- High-precision RAG and reading functions: Cross-references internal documents such as PDF, Word, Excel, PowerPoint, DocuWorks, and external data sources such as Box and SharePoint
- Rich built-in tools: Deep Research, PowerPoint generation, Excel generation, Excel/CSV analysis, image generation, etc.
- Flexible user interface: ChatGPT-like chat