Retrieval-Augmented Generation (RAG) is the agent’s knowledge layer. It allows the agent to fetch relevant information from external sources (databases, documents, APIs, vector stores) before generating a response.
Think of RAG as the agent’s memory and research system — it ensures answers are grounded in real, up-to-date, and company-specific data rather than just the model’s general training.
In the Agent Workspace you can:
- Upload documents, connect site content, or point to structured data.
- The agent uses this information to provide accurate, context-rich
answers.
- This is ideal for support, product documentation, or internal knowledge agents.
