Resource Reference

Represents a reference to an external resource that an AI Assistant can use during a chat conversation.

Remarks

A Resource Reference is a foundational concept for enabling Retrieval-Augmented Generation (RAG) in conversational AI systems. RAG is a technique that combines the generative capabilities of large language models (LLMs) with the precision and breadth of external knowledge sources. By augmenting the model's responses with information retrieved from trusted resources, RAG improves the accuracy, context, and trustworthiness of AI-generated answers.

Theory Behind RAG:

• RAG works by allowing the AI Assistant to retrieve relevant information from external resources at runtime, rather than relying solely on its pre-trained knowledge.
• When a user asks a question, the assistant can search, query, or invoke external resources—such as search indexes, databases, web services, or knowledge bases—to gather up-to-date or domain-specific information.
• The retrieved content is then incorporated into the assistant's response, resulting in answers that are more accurate, current, and context-aware.

Role of Resource References:

• A Resource Reference acts as a bridge between the AI Assistant and external resources. It provides the configuration and metadata needed for the assistant to discover, describe, and interact with these resources during a conversation.
• Each resource reference defines how the assistant can access a particular type of resource, such as a Search Index, Database Table, Web Service, or Bing Search.
• By exposing these resources in a structured and discoverable way, resource references enable the assistant to select the most appropriate tool or data source for a given user query.

Types of Resources:

• Search Indexes: Allow the assistant to perform semantic or keyword-based retrieval over large collections of documents, supporting question answering and knowledge discovery.
• Database Tables: Enable direct querying of structured data for reporting, analytics, or fact-based responses.
• Web Services: Provide integration with external APIs, allowing the assistant to fetch live data, trigger workflows, or interact with third-party systems.
• Web Search (e.g., Bing): Grants access to real-time information from the internet, supporting fact-checking and up-to-date answers.

How Resource References Facilitate RAG Conversations:

• When a user asks a question, the AI Assistant evaluates which resource references are available and relevant.
• The assistant can retrieve supporting information from one or more resources, synthesize the results, and generate a response that is both informed and contextually appropriate.
• This approach ensures that answers are not limited by the assistant's training data, but can leverage the full breadth of organizational and external knowledge.

By configuring Resource References, you empower the AI Assistant to deliver more reliable, explainable, and actionable responses—making RAG a practical and powerful enhancement for enterprise AI conversations.

Properties

Name	Type		Description
Resource Name	String	►	The name used to identify this resource. The 'Resource Name' is shown in the user interface and is used to reference this resource in configuration and during chat interactions. If an 'Alias' is provided, it will be used as the display name; otherwise, the default 'Name' is used.
Description	String	►	A description of the resource to help the AI Assistant understand its purpose and relationship to other resources. EditorText Property Editor Provide a clear and concise explanation of what this resource is for, what information or capabilities it provides, and how it should be used in relation to other available resources. This description is used by the AI Assistant to determine when and how to utilize the resource during a conversation.
Alias	String	►	An optional alias for the resource, used to resolve naming conflicts. If two resources would otherwise have the same 'Resource Name', you can specify an 'Alias' to ensure each resource can be uniquely identified. The alias will be used as the display name in the user interface and in chat interactions.

Derived Types

There are 4 implementations of Resource Reference.

Bing Search	Exposes Bing's web search service as a resource that an AI Assistant can use for real-time web search, fact-checking, and knowledge augmentation.
Database Table	Exposes a database table as a resource that an AI Assistant can use to answer questions or provide data-driven insights.
Search Index	Exposes a search index containing documents as a resource that an AI Assistant can use for semantic and standard search, document retrieval, and knowledge discovery.
Web Service	Exposes a web service as a resource that an AI Assistant can use for dynamic data retrieval, integration, or automation.

Used By

AI Assistant