Learn the Basics of Grounding
Learning Objectives
After completing this unit, you’ll be able to:
- Understand the concept of grounding an agent and why it helps.
- Clarify how grounding and retrieval augmented generation (RAG) are related.
- Describe the types of data that can be used as data sources for Agentforce.
Before You Begin
We know you’re eager to learn about agents and grounding them with data as we roll out new generative artificial intelligence (AI) tools. Before you begin, consider completing this recommended content.
The Einstein Generative AI Glossary of Terms also covers many of the terms used in this badge, such as large language models (LLM), prompts, grounding, hallucinations, toxic language, and more. Learn more about LLMs in the Large Language Models Trailhead module.
What’s Grounding?
In general, grounding is the idea of infusing an LLM prompt with the information that you want the LLM to consider when processing a request. Grounding data sources can include structured data, like Excel spreadsheets and CRM data, as well as unstructured data, like PDFs, chat logs, email messages, and blog posts.
The goal of grounding is to improve the accuracy and relevance of LLM responses. LLMs, which are frequently used with agents, are trained on general rather than context-specific knowledge or proprietary information. Adding pertinent, domain-specific knowledge and contextual information from reliable data sources enhances LLM results and increases trust in AI solutions.
Grounding agents with verifiable data sources leads to better decision-making and more effective actions. Agents succeed better when they’re provided with the most up-to-date, accurate, and relevant information.
Structured and Unstructured Data
The information used to ground agents and LLMs can be from structured and unstructured data.
Structured data is organized in a predefined format with known metadata. It’s easy to search, analyze, and integrate structured data with agents. Examples include data from Salesforce objects, like Accounts, Contacts, and Cases, or Data Model Objects (DMOs).
Unstructured data lacks a predefined format with unknown metadata. Examples include emails, chat logs, social media posts, or documents. While more challenging to process, unstructured data can provide valuable insights. Unstructured data requires preprocessing to prepare and optimize the knowledge for retrieval. To connect this data to Agentforce, you might use natural language processing (NLP) tools or data extraction services to convert it to a structured format.
Grounding with Retrieval Augmented Generation (RAG)
RAG is a form of grounding that retrieves knowledge from unstructured data sources. In this approach, the LLM prompt is enriched with relevant, current information from a trusted source, such as a collection of documents. For example, if a customer asks about a product feature, RAG gets the latest feature details from your knowledge base and adds this to the prompt from which the LLM generates a response.
To make grounding with data practical and effective, you need to understand some ways to use it in Agentforce. You explore those two ideas next.