SSSUG3T_1.0.0 - Documentation Index

Table of Contents

• Welcome
• Release notes
  • What's new
  • Known issues and limitations
• Getting started with IBM Content Assistant Client Managed Software
  • Product overview
    • Use cases and scenarios for IBM Content Assistant
• Preparing to install
  • Downloading and extracting the IBM Content Assistant Client Managed Software package
  • Preparing your system to use IBM Content Assistant Client Managed Software
  • AI provider integration
    • Option 1: Creating a watsonx.ai account for the IBM Content Assistant Client Managed Software
      • Configuring a service ID for your watsonx.ai account
    • Option 2: Installing IBM watsonx.ai Lightweight Engine
  • Setting up your OpenSearch cluster
    • Cluster sizing considerations
      • Estimating the size of the OpenSearch index for each object store
      • Sizing your OpenSearch cluster
    • Installing OpenSearch cluster
      • Option 1: Installing a standalone OpenSearch cluster
      • Option 2: Installing OpenSearch cluster from CP4BA deployment using a script
        • (Optional) Installing OpenSearch cluster from CP4BA deployment manually
    • Configuring OpenSearch cluster to access FNCM GenAI Connector Service
      • Configuring cluster security
      • Configuring security for CP4BA OpenSearch
        • Configuring security using Kubernetes job
        • Configuring security using curl commands
• Installing and setting up the Content Assistant
  • Installing and deploying the FNCM GenAI Connector Service
    • System requirements
    • Installing the FNCM GenAI Connector Service
      • Preparing your Kubernetes cluster
        • Preparing a client to connect to the Kubernetes cluster
        • Downloading the ibm-content-assistant-client-managed-software repository
        • Preparing a namespace for the operator
        • Getting access to images from the public IBM entitled registry
        • Preparing your cluster for an air-gapped (offline) deployment on OCP
          • Setting up a host to mirror images to a private registry
          • Setting up a private registry
          • Downloading CASE packages and mirroring images in an airgap environment
            • (Optional) Downloading the CASE package manually
            • (Optional) Mirroring images to the private registry manually
              • Mirroring catalogs to a private registry using oc mirror
                • Mirroring catalogs to a private registry with a bastion server
                • Mirroring catalogs to a private registry with a portable compute or a storage device
        • Preparing your cluster for an airgapped (offline) deployment on CNCF
          • Getting access to images from a local image registry
      • Installing the Content Assistant operator
        • Installing the operator using the OpenShift console
        • Installing the operator by running a script
        • Installing the operator manually
      • Uninstalling and removing components
        • Uninstalling from the OCP console
        • Uninstalling by running a script
        • Uninstalling manually
    • Configuring the FNCM GenAI Connector Service
      • Generating simple custom resource and deployment files
      • Custom resource reference
        • Shared configuration
        • Vector database parameters
        • Content Assistant parameters
    • Deploying the FNCM GenAI Connector Service
      • Applying the custom resource manually
      • Validating your deployment
      • Accessing your deployment
      • Modifying your deployment
  • Setting up IBM Content Assistant
    • Preparing Content Platform Engine
      • Deploying the Content Assistant jar file
      • Importing the Content Assistant certificate into containerized Content Platform Engine deployments
      • Importing the Content Assistant root certificate into traditional Content Platform Engine deployments
      • Installing the Content Assistant add-on for an object store
      • Enabling persistent text extraction
        • Installing the persistent text extraction add-on
        • Configuring text extraction for document classes
        • Reviewing the storage configuration for the Annotation class
      • Registering an object store with the FNCM GenAI Connector Service
      • Creating a document class subscription for vector indexing
      • Creating a document class subscription for watsonx summary
    • (Optional) Preparing IBM Content Navigator
      • Registering the IBM Content Assistant plug-in for Content Navigator
    • Removing Content Assistant
      • Disabling vector indexing for an object store
      • Deleting subscriptions related to the Content Assistant service
        • Deleting persistent text extraction subscriptions (CPE V5.5.12 and earlier)
        • Deleting vector indexing subscriptions
        • Deleting watsonx summary subscriptions
      • Disabling text extraction for document classes (CPE V5.6 and later)
      • Disabling the Content Assistant for a desktop
      • Disabling the Content Assistant on Content Navigator
      • Deleting the Content Assistant plug-in on Content Navigator
• Upgrading
  • Preparing to upgrade the FNCM GenAI Connector Service
    • Upgrading the operator version and your components
      • Upgrading with an existing operator in the OpenShift console
      • Upgrading the operator by running a script
      • Upgrading the operator manually
    • Verifying the upgrade
  • Upgrading the Content Assistant artifacts for Content Platform Engine
  • Upgrading the IBM Content Assistant plug-in for Content Navigator
• Administering
  • Configuring IBM Content Assistant
    • Configuring security
      • Managing certificates
        • Changing the default root CA signer certificate
        • Creating secure endpoints for external services
        • Connecting endpoints to external services over TLS
          • Importing the certificate of an external service
          • Exporting the operator root CA key and importing it into an external service
      • Configuring cluster security
      • Configuring cluster network policy
        • Retrieving operator-generated network policy templates by using a script
        • Retrieving operator-generated network policy templates manually
      • Verifying container image integrity
    • Enabling the Content Assistant plug-in
    • Configuring Ask Your Repository for a desktop
    • (Optional) Configuring commonly asked questions for the Content Assistant
    • (Optional) Configuring the default multi-document query mode for a desktop
    • Configuring document comparison
      • Configuring the Document Compare menu action
      • Configuring the View Version menu action
    • Configuring the FNCM GenAI Connector Service parameters for an object store
      • Changing the Content Assistant large language model for an object store
      • Changing the Content Assistant embedding model for an object store
      • Configuring Content Assistant to use enhanced annotations
      • Changing the Content Assistant index status for an object store
      • (Optional) Managing the default prompt templates
        • Synchronizing the prompt template values for an object store
        • Editing the default prompt templates
        • Creating a summary prompt template for a document class
        • Specifying the prompt template by using the API
      • Setting the maximum summary words
      • Configuring advanced options
        • Adjusting the relevancy score filter level
        • Configuring default chunk settings for search queries
        • Enabling adaptive RAG mode for Ask your repository search
    • Configuring AI provider settings
    • Configuring the FNCM GenAI Connector Service parameters using the command-line
      • Installing the IBM Content Assistant CLI tool
      • IBM Content Assistant command-line tool reference
        • How to read the syntax diagrams
        • config command
        • login command
        • list command
        • update command
        • delete command
        • metric command
        • index template command
  • Tuning IBM Content Assistant
    • (Optional) Running a custom text extraction job sweep for existing documents
    • (Optional) Running a custom job sweep for summarizing existing documents
    • (Optional) Running a custom job sweep for vector indexing existing documents
    • (Optional) Running a custom job sweep for de-indexing existing documents from the vector database
    • Reindexing document content
    • Restoring the vector index for document content
    • Adjusting the number of queue sweep workers
• Using the Content Assistant
  • Asking questions about a single document
  • Asking questions about multiple documents
  • Asking questions about content across the repository
  • Accessing the watsonx summary for a document
  • Accessing the watsonx ad hoc summary for your document selection
  • Comparing two documents
  • Comparing two versions of a document
• Developing generative AI applications
  • GenAI query classes
    • GenaiBaseQuery
    • GenaiVectorQuery
    • GenaiDocumentQuery
    • GenaiMultiDocumentQuery
    • GenaiAdhocSummary
    • GenaiDocumentComparison
  • Handling GenAI query objects
  • Content Engine Java API samples
  • Content Engine GraphQL API sample mutations and queries
• Troubleshooting and support
  • Troubleshooting the FNCM GenAI Connector Service
  • Trouble summarizing one or multiple documents
  • Trouble with questions that involve a mathematical calculation
  • Trouble indexing and summarizing certain documents