What Usage Data Does Harvey Collect?

What is Usage Data?

Usage data is information about how users interact with and use Harvey. This data provides valuable insights to both Harvey and its customers.

Types of Usage Data Collected

Harvey limits the amount of information it collects, gathering only what is required to generate these insights. It collects usage data such as:

• The frequency with which a user logs into Harvey and the number of prompts they submit.
• When documents are uploaded to Harvey any related metadata including, when it is uploaded, where it is uploaded from (such as through SharePoint or Google Drive), and the document type.
• Whether a user adds a knowledge source to their prompt.
• Where a workflow or playbook is used.
• The specific features or functions a user interacts with.
• How users navigate and click through the Harvey platform.

Custom Workflow Data

For custom workflows, Harvey collects additional information, such as the name and description of each workflow. This data helps Harvey understand how customers are using custom workflows, identify trends, and determine which features or developments would be most beneficial for users.

How Harvey Uses this Data

Harvey prioritizes transparency by providing customers with access to their own usage data. Customers can view detailed insights about their personnel’s engagement with Harvey through the Analytics dashboard or, on request, an API, which is available in the Admin settings.

We also collect usage data to enhance the user experience and inform the development of new features and updates. By analyzing high-level trends and engagement patterns, Harvey can make data-driven decisions on how to improve its services.

What’s Included in the Analytics Dashboard

The Analytics dashboard illustrates key metrics, like the number of unique users or query volumes, that can be viewed over customizable time ranges. Data can be segmented by product feature and user profile attributes for detailed insights. Additionally, Admins can see an aggregated count of user queries across different product features, enabling identification of usage patterns and popular features.