NEW Feature: Smart Categories and Document Preview

By Axel Kirstetter, Product Marketing and Dan Tschida, Product Management

As we do discovery with our administrators, there is one consistent piece of feedback: “I need Datasite to help me spend my time on higher value tasks”. With smart categories and document preview we give you the tech to do just that. Allow our natural language processing algorithm to suggest categories of newly loaded documents and preview them before you confirm their belonging. 

It's All About the Index

With our experience of powering 9,000+ deals per annum we know there is no single way for launching a data room. Some administrators spend a lot of time getting the index right. Others rush to launch the project room without a final structure knowing the Datasite Diligence search capability will help early bidders.

With smart categories for Datasite Diligence our powerful AI engine – trained on 3m+ documents and an M&A vocabulary of 300k+ words – automatically suggests a category for a document.

How It Works

1. Load your documents into Datasite via drag and drop for instance
2. Have the system suggest the category
3. Confirm the category by either
4. Approving the AI identified category

OR selecting from our list of over 100 categories

Need to quickly find all documents relating to ‘Benefits’ irrespective of their index location? Simply filter for what you are looking for and Datasite will bring up all documents that fit that category.

Additional Value

But wait. It gets better. We understand that ‘executive compensation’ could be categorized as HR, Contracts or Corporate Governance. To give our admins more control over category selection, hover over the document name which will give a preview of its contents. No need to downloads the document to know what it contains. 

Saving Admins Time

At Datasite we aim to save admins time with user experience improvements and use of modern technology. The former is in the form of fewer clicks, like the ability to preview documents. The latter with the use of machine learning.