Table of Contents

Capture, generate and use data with AI Agents

Pau Sanchez Updated by Pau Sanchez

AI Agents are not just useful tools to chat about frequently asked questions - they can do much more. With AI Agents, you can capture information (useful for cases like lead generation) and also generate data based on conversations.

Capture data through Questions

1. Set Up Card

Go to the Get User Information section

You'll see quick examples like name and email already available

To add custom fields for other type of questions, click Use Field

Select where you want to store the information, in which field (e.g., create a field called "location")

Specify to the agent the which information must be stored in this field (e.g, "The city of the lead")

2. Set Up Instructions

To make the agent ask for information, you must specify this in the main (prompt) Instructions:

If you selected email and email, add specify it in the instructions section

If you added custom data to be captured like "location", add it in the instruction sections

Important: Be as specific and detailed as possible about how the information must be captured

  • If you want budget ranges like 0-100, 101-200, etc., provide examples
  • Show how to handle responses: if user says "22", it should be stored as "0 to 100"
  • Prepare for edge cases when users don't know their budget
  • Offer options so the agent can be consistent in capturing this information

View Analyze Section Data

To see the captured information, you have two options:

  • Main option: Go to the Analyze section where you can see all captured information

  • If agent is part of a flow: You must go to the Analyze section of the chatbot flow where the conversation starts, not the agent itself

Filter for the time period you're interested in to view your data.

Generate data out of the conversation interactions

Beyond capturing direct answers, configure your agent to analyze and generate insights from conversations.

Example Conversation

The agent can create structured data based on conversation context, user responses, and behavioral patterns.

View Analyze

Use the same process as above - check the appropriate Analyze section based on whether your agent is standalone or part of a larger flow.

Use data outside the AI Agent

Once you capture information, you can use it throughout the chatbot experience. Make sure to use the same field names and data types for consistency.

Use it, for CRMs, or Notifications, or to control the flow...

Give Context to Improve Behaviuor

Connect Bot

When coming from a chatbot flow where you've already captured information in previous blocks, you can enhance the AI Agent experience.

Ask Name Block + Variable Usage

If you capture the user's name in an earlier block (like a form), you can pass this as a variable to the agent.

Set up instructions

In your agent instructions, write: "You are speaking with @name"

This allows the agent to:

  • Personalize the experience immediately
  • Refer directly to the user without asking for their name again
  • Create a much richer, more natural conversation flow

Was this article helpful?

How to create custom Instructions for your Landbot AI Agent with AI (ChatGPT, Claude...)

Contact