Exploring Now Assist in Document Intelligence

  • Release version: Zurich
  • Updated July 31, 2025
  • 3 minutes to read
  • Summarize
    Summarized using AI
    This content was generated using new OpenAI-powered functionality. Results are provided on an as is basis and are not guaranteed to be accurate or complete.

    Summary of Exploring Now Assist in Document Intelligence

    Now Assist in Document Intelligence leverages generative AI to extract key information from documents and images to streamline workflows within the ServiceNow platform. It provides capabilities such as extracting fields, tables, and answers to predefined questions, enabling agents to review and utilize this data effectively in their tasks.

    Show full answer Show less

    The solution includes AI-powered features like document Q&A, attachment summarization, and document chat, which enhance agent productivity by providing predictive answers and summaries based on document content. Specialized skills for contract metadata extraction and invoice data extraction are also available, supporting use cases in contract management and accounts payable operations.

    Key Features

    • Information Extraction: Extract structured data—fields, tables, and answers—from documents or images, with options for agent review in the Document Intelligence workspace.
    • Document Q&A: AI provides predictive answers to predefined questions, reducing manual review time.
    • Document and Visual Insights AI Agent: Generates contextual responses based on user input and document attachments, aiding task completion.
    • Attachment Summarization: Summarizes attachment content in IT Service Management (ITSM) records to quickly inform agents.
    • Document Chat: Enables chat-based responses derived from document content within Virtual Agent interactions.
    • Invoice Data Extraction: Automates extraction and processing of invoice details from PDFs for Accounts Payable Operations (APO).
    • ESG Information Extraction: Extracts environmental, social, and governance data from invoices to support sustainability initiatives.
    • Now Assist Skills Management: Skills like extraction and multimodal chat are enabled and managed through the Now Assist Admin Console to tailor AI capabilities to specific workflow needs.

    How It Works

    The workflow begins with an administrator activating a relevant AI skill and configuring the use case. During process execution, documents are uploaded and processed automatically by Now Assist, which predicts relevant data fields. If full automation is not enabled, these predictions are reviewed and corrected by live agents within the Document Intelligence workspace before the workflow continues.

    Benefits

    • Guided Setup: Simplifies configuration by guiding admins through defining use cases, accelerating deployment.
    • Increased Efficiency: Automates extraction of structured data from documents, reducing manual data entry and review times.
    • Improved Accuracy: Allows agents to validate AI predictions, ensuring high-quality data for downstream workflows.
    • Expanded Use Cases: Supports diverse document types and scenarios, including contracts, invoices, and ESG reporting, enhancing operational capabilities.

    With Now Assist in Document Intelligence, you can use generative AI to extract key information from your documents and images for use in your workflows.

    Now Assist in Document Intelligence overview

    Now Assist in Document Intelligence makes the following generative AI capabilities available:
    • Extract information from documents. Extract information in the form of fields, tables, and answers to predefined questions from documents and image files. The information can be reviewed by agents in the Document Intelligence workspace, stored in mapped fields, and used as defined in the workflow.
    • Document Q&A: Agents can save time when reviewing documents by using the predictive capabilities of generative AI to provide answers to predefined questions.
    • Document and visual insights AI agent. The AI agent gathers context from user input and document or image attachments, generates the requested information based on the content, and provides the information along with any relevant task details. For more information on AI agents, see Document and visual insights AI agent.
    • Contract metadata extraction and contract analysis. The skills for contract metadata extraction and contract analysis are only available with the Now Assist in Contract Management application. For more information, see Now Assist in Contract Management.
    • Attachment summarization. Agents can view a summary of attachment content along with the record summary in ITSM. Attachment summarization is available in Now Assist for ITSM. For more information, see Customize a Now Assist for IT Service Management (ITSM) skill.
    • Document chat. Agents can receive chat responses based on document content. Document chat is available in Now Assist for Virtual Agent. For more information, see Upload documents in a chat.
    • Invoice data extraction. Extract invoice details from PDFs submitted through email or supplier portal and automatically process the digitized invoice within Accounts Payable Operations (APO). For more information, see Now Assist for Accounts Payable Operations (APO).
    • ESG information extraction. Extract ESG-related information from invoices to support Operational Sustainability Management (ESG) initiatives and regulatory compliance. For more information, see Filter citations documents for Operational Sustainability Management (formerly ESG Management).

    Now Assist in Document Intelligence skills

    Now Assist in Document Intelligence skills are enabled in the Platform workflow on the Now Assist Admin Console. For more information on activating the skills, see Activate a Now Assist in Document Intelligence skill.
    Extract information from documents
    The extract information from documents skill allows you to use Now Assist predictions to extract fields and tables and find answers to predefined questions from a document or image. Agents can review and edit predictions in the Document Intelligence workspace. For more information, see Review extracted information in the Document Intelligence workspace.
    Multimodal chat

    The multimodal chat skill is used to enable chat responses about the content of uploaded documents and images. It is only used on the server side by the document and visual insights AI agent and by the question answering capability in Now Assist for Virtual Agent. This skill doesn’t require configuration in the Now Assist Admin console.

    For more information on these capabilities, see Document and visual insights AI agent and Upload documents in a chat.

    Now Assist in Document Intelligence workflow

    The following diagram shows how the Now Assist in Document Intelligence skills are set up and used to process documents.

    Figure 1. Now Assist in Document Intelligence flow
    Diagram showing the activity flow in Now Assist in Document Intelligence.
    In this workflow:
    • An admin activates a skill and sets up a use case for it.
    • A workflow integration creates a document task as part of its flow.
    • A document is uploaded for processing in a document task.
    • Now Assist processes the document and makes predictions based on the fields defined in the use case.
    • If the use case is not set to full automation, the task is sent to a live agent for review.
    • The agent provides input to validate or correct the values predicted by Now Assist.
    • The task is completed and the integrated workflow proceeds as defined.

    Now Assist in Document Intelligence benefits

    Now Assist in Document Intelligence provides the following benefits.
    Benefit Feature User
    Start fast with a guided set up of your use cases to identify the information you want to get from your documents.

    Set up document intelligence use cases

    DocIntel Admin [sn_docintel.admin]

    DocIntel Manager [sn_docintel.manager]

    Accelerate the extraction of information from documents in the form of fields, tables, and answers to predefined questions, and turn it into structured data in the platform.

    Review extracted information in the Document Intelligence workspace

    DocIntel Extraction Agent [sn_docintel.extraction_agent]