> ## Documentation Index
> Fetch the complete documentation index at: https://docs.datachecks.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Agents Overview

> Datachecks uses four AI agents to automate discovery, assessment, validation, and translation across your migration.

Datachecks is built around four AI agents, each responsible for a distinct phase of the data migration lifecycle. The agents are powered by [LangGraph](https://langchain-ai.github.io/langgraph/) workflows and run as background jobs you can monitor in real time.

## The four agents

<CardGroup cols={2}>
  <Card title="Asset Discovery" icon="magnifying-glass" href="/agents/asset-discovery">
    Ask natural language questions about your data schemas, tables, columns, and relationships. The agent queries your registered assets and generates SQL on demand.
  </Card>

  <Card title="Migration Assessment" icon="chart-bar" href="/agents/migration-assessment">
    Analyze stored procedures, views, and translation complexity in your source database. Get a structured readiness report before committing to a migration.
  </Card>

  <Card title="Data Validation" icon="shield-check" href="/agents/data-validation">
    Describe what you want to validate in plain language. The agent extracts the target table and validation type and configures the validation for you.
  </Card>

  <Card title="Translation" icon="code" href="/agents/translation">
    Translate SQL queries from your source database dialect to the target platform. The agent classifies each query, translates it, and verifies correctness by running a comparison.
  </Card>
</CardGroup>

## How agents fit into the migration workflow

The agents are designed to work sequentially, but you can run them independently:

```
Asset Discovery
      ↓
  Understand what you have
      ↓
Migration Assessment
      ↓
  Know what will be complex
      ↓
Translation
      ↓
  Convert SQL logic to target dialect
      ↓
Data Validation
      ↓
  Confirm data integrity after migration
```

## Where to find agents

Navigate to the **Agents** tab in the left sidebar to access the agent interface. Each agent run is tracked as a job — go to **Jobs** to monitor progress and view results.