AAirbyteHelp Center

Guide

Query your Agent

4 stepsAgent View (JSON)

Overview

Airbyte's Agent Engine enables you to interact with your connected data sources through conversational AI. Instead of writing SQL queries or navigating complex interfaces, you can ask your agent questions in plain language about your data—such as 'How many customers signed up last week?' or 'Show me the top performing products'—and receive immediate answers. This workflow is essential when you want to perform ad hoc data exploration, get quick insights from your connections, or ask your agent to perform read operations across multiple authenticated sources without switching between tools.

Before you begin

  • You must have an active Airbyte account and be logged into app.airbyte.ai.
  • At least one data source connector must be enabled and authenticated in your Airbyte Agent Engine.
  • You should have already set up at least one connection between a source (e.g., database, SaaS platform) and your agent.
  • Access to the Agent Engine chat interface (confirm this is enabled in your workspace settings).

Step by step

1
ClickChat with your Agent

Navigate to the chat interface by clicking the 'Chat with your Agent' link in the Airbyte application. This opens the conversational query interface where you can interact with your Agent using natural language.

Tip. The 'Chat with your Agent' link is typically located in the main navigation or sidebar. Bookmarking this page will let you return to it quickly for future queries.
Step 1
2
ClickAsk about data from connected sources...

Click on the text input box labeled 'Ask about data from connected sources...' to focus the input field and prepare to type your question.

Tip. The placeholder text 'Ask about data from connected sources...' remains visible until you begin typing. You can also press Tab to navigate to this field if you prefer keyboard shortcuts.
Step 2
3
TypeAsk about data from connected sources...

Type your question in natural language. For example, ask the agent to fetch, search, or summarize data from your connected sources—such as 'Show me all customers from New York' or 'What is the average order value this month?'

Tip. Be as specific as possible with your question, including relevant table names, field names, date ranges, or filters. This helps the agent understand your intent and return more accurate results. The agent supports read operations and can search across all authenticated sources you have connected.
4
Keyboard

Press Enter to submit your question to the Agent. The agent will process your query and return results or a response based on the data available in your connected sources.

Tip. After pressing Enter, monitor the chat window for the agent's response. Most queries resolve within a few seconds. If you want to ask a follow-up question, simply click the input box again and repeat steps 2–4.
Step 4

Confirm it worked

  1. 1You see the chat interface open with a message input box labeled 'Ask about data from connected sources...'
  2. 2After typing your question and pressing Enter, the agent responds with a message or data result within the chat window.
  3. 3The agent's response references the data from your connected sources (e.g., mentions specific tables, records, or values you know exist).
  4. 4No error messages appear; the chat remains interactive and ready for follow-up questions.

Common issues

Keep reading

Your First Agent with query() | Agent Factory

Your First Agent with query() | Agent Factory On this page Teach Me Sign in to access Teach Me mode Ask Sign in to ask questions Copy as MarkdownCtrl+⇧+C Updated Mar 16, 2026 You're about to w

agentfactory.panaversity.org

Query Agent | Weaviate Documentation

Query Agent | Weaviate Documentation Go to documentation: ⌘U ✕ Weaviate Database Develop AI applications using Weaviate's APIs and tools Deploy Deploy, configure, and maintain Weaviate Database

weaviate.io

Introducing the Weaviate Query Agent | Weaviate

Introducing the Weaviate Query Agent | Weaviate ← Back to Blogs We’re incredibly excited to announce that we’ve released a brand new service for our Serverless Weaviate Cloud users (including free S

weaviate.io