---
title: 3. Path Setup
slug: 3-path-setup
docTags: 
createdAt: 2026-03-23T19:57:31.784Z
---

After completing **Data Training**, the next step is to configure the **Path Setup**.

Paths define how the agent retrieves and uses the trained data during conversations. By configuring paths, you allow the agent to fetch information from trained sources and generate responses based on that data.

In this step, you will create paths using **JSON node APIs**, which connect the trained data to the agent’s workflow.

The following path configurations are available:

**RAG Path Setup**

- Configure a path to retrieve information from trained documents using **RAG (Retrieval-Augmented Generation)**.

::embed[]{url="https://youtu.be/0IavnNbXTpg?si=j8XbbjnWWdY_JglP"}

**Mongo Path Setup**

- Configure a path to fetch dynamic data from a **MongoDB database**.

Once the paths are configured, the agent can access the required data during conversations.

::embed[]{url="https://youtu.be/ep12K_vqQfI?si=7W018NsA5_NbOpF0"}

📚 Next Steps

Continue to [4. Workflow Setup](docId\:uUhjiZq8X_4pijsSvst9Q) to connect the configured paths with the agent’s conversation flow.