News
Box and Google Cloud have extended their existing partnership in response to the growing use of generative AI in the workplace. The result is the integration of Box and Google’s Vertex AI platform, enabling customers of both to build new generative AI features that help customers process and analyze the data and content they have stored in Box Content Cloud.
The Vertex AI Advantage
In a statement about the extended partnership, Thomas Kurian, CEO of Google Cloud, said the use of Vertex AI will make it possible for users to create insights and processes from content stored in Box and, specifically, will help process data faster to create personalized user experiences as well as providing them with intelligent search. All of this will feed into the development of LLMs.
Vertex AI is a machine learning (ML) platform that lets organizations train and deploy ML models and AI applications as well as customize large language models (LLMs) for use in AI-powered applications.
It combines data engineering, data science and ML engineering workflows, so teams can collaborate using a common toolset and scale applications using Google Cloud. The announcement follows earlier moves to integrate Google Cloud’s advanced large-language models (LLMs) into Box AI.
Vertex AI offers organizations model training and deployment by offering:
- AutoML: Organizations can train tabular, image, text or video data without writing code.
- Custom training: Users have complete control over the training process, including using ML framework and writing code for training.
- Model Garden: Offers testing, customization and deployment of Vertex AI and other open-source (OSS) models and assets.
- Generative AI: Offers access to Google's large generative AI models for multiple modalities (text, code, images, speech).
The feature that will be offered first as an API aims to improve data management by pulling content from their documents that will include matching metadata fields and attributes within a file. There are a few advantages to this. With it, users can:
- Automatically classify and label documents at a massive scale to pull the main insights that the documents provide.
- Provide metadata templates that can be used to extract data for custom use cases and automation recognize and tag content in images, or — probably most important in the current security-conscious climate — categorize personal identifiable information.
- Identify and protect key information including document version history, authorship and timestamps.
Logical Partnership
The extension of the partnership is a logical move, said Box CTO Ben Kus in an email interview. Box, he said, has been using Vertex AI machine learning ops and other machine learning-related capabilities for some time.
“We consider it one of the premier solutions in terms of both capabilities as well as security, compliance and privacy, which we believe is critical when talking about applying AI to enterprise content,” he said.
Box, he added, is platform-neutral regarding AI vendors, which means it uses the best models for different use-cases. Google has been pushing forward with the development of its AI technologies and he views its models as best-in-class for solutions like extracting metadata and other generative AI capabilities. He also stressed the importance of security in the partnership.
“The security, compliance and privacy from a vendor like Google Cloud is especially important when applying AI and to align with Box's Box AI Principles, particularly our commitment to transparency," he said.
There is another advantage for Box users too. He pointed that the new solutions are provided by Box, so organizations don’t need to worry about managing infrastructure. Box manages the use of the AI models that are provided from Vertex AI.
“Box takes an AI neutral approach and is transparent about which models are being used and for what but takes care of managing the infrastructure, so customers don't have to."
Box AI with Vertex
Kus points out that Box AI was only released in May, with two initial capabilities: Document Q&A in preview and content generation in Box Notes.
Building on this, at BoxWorks on Oct 11, Box announced multi-document search and summarization based on the documents users organize in a “Hub” so users can find answers to critical questions in quickly, automatically summarize vast amounts of information, and create new content.
By adding Vertex AI it is adding metadata capabilities to Box AI, which will automatically extract metadata on documents using Generative AI.
"We see this as a foundational capability that we will continue to enhance and will help customers understand and extract more insights from their large amounts of unstructured data that is stored in Box,” he added.
Google has been working on its large language models (LLMs) for decades and has some of the most advanced generative AI technology on the market, even if the recent figures for the last quarter show it is struggling to gain traction with large enterprises.
Even still, Vertex AI includes a broad set of offerings including generative AI PALM models, embeddings models, fast approximate-nearest-neighbour lookup to power Retrieval-Augmented-Generation, and more.
Kus said they have opted for Vertex because after testing it was and found to perform very well for Box’s deep content use cases.
Conference
Sep
16
HR Tech Conference Las Vegas 2025
Register
Conference
Oct
27
Gartner HR Symposium/Xpo Orlando 2025
Register
The expanded partnership also builds on existing integrations with Google Workspace, which enables Box customers create, collaborate, and save content in Google Docs, Sheets, or Slides from the Box Content Cloud platform. The Box for Google Workspace add-on also enables collaboration across Google Workspace, including Gmail, Google Drive, and Google Calendar.
Box is already available on the Google Cloud Marketplace, but the extraction API functionality is expected to begin rolling out to Box Enterprise Plus plan users in beta next year.
Have a tip to share with our editorial team? Drop us a line: