SDK

Low-Code

Workflow

DWS API

Developer guides

Welcome to the AI Assistant developer guides. These guides show how to set up AI Assistant to provide AI functionality in Nutrient Web SDK.

A quick infrastructure primer

To have a fully functional AI setup, you’ll need:

  • Nutrient Web SDK — A document viewer in the browser that also exposes the user interface (UI) for the AI features

  • AI Assistant — A service to process the AI request and process documents

  • Document Engine (optional) — A service that stores and serves documents to AI Assistant and Nutrient Web SDK

If you’re brand new to Nutrient, take a look at our getting started guides to quickly add AI functionality to a new web project.

Concepts

There are two important concepts to be aware of when integrating AI Assistant:

  • documentId — A unique identifier for your document. When using Document Engine, this is returned upon document upload (see the API reference).

  • sessionId — An alphanumeric unique ID specified by you to allow recall of a chat history if a user navigates away.

These two concepts are used in the configuration object passed to Nutrient Web SDK, and they can be used to limit user access to specific documents and sessions via the claims of a JSON Web Token (JWT).

There’s one additional concept, which is loosely linked with sessionIds, but this is optional, depending on your requirements:

  • userId (optional) — Although it’s not required, this is an ID you specify that makes it possible to implement usage limits via claims in a JWT, or to purge data specifically related to a single user via the server API.

Guides

Docker configuration
How to customize AI Assistant’s options

AI model configuration
How to specify the models AI Assistant uses

Deployment
How to deploy AI Assistant

Viewer integration
Learn how to connect Nutrient Web SDK with your AI Assistant instance

Data management
Take control of your users’ data with the server API