• INVENTORY
  • SEARCH
  • AUDIT
  • HUMAN

How to manage extraction models with your AI

Connect Cradl AI to your agent to treat your custom models like a searchable database. You can query model performance and metadata without leaving your chat interface.

Ask AI about this page

Short answer

How can I manage my extraction models using AI?

Your AI uses Cradl AI to pull lists of your models, find specific ones by name, and grab detailed accuracy metrics. This lets you audit model performance through natural language queries.

Model management outcomes

Where your model data lands.

These are the specific ways your AI interacts with your Cradl AI data.

INVENTORY

Model lists

The AI pulls a full directory of your custom-trained models. You get a clear view of everything currently in your library.

SEARCH

Targeted lookups

Find a specific model instantly by its name. Your AI handles the filtering so you don't have to scroll through lists.

AUDIT

Performance checks

Retrieve specific accuracy metrics and metadata for any model. This gives you the raw data needed to judge model quality.

HUMAN

Manual updates

The AI only reads and retrieves data. You still need to use the Cradl AI dashboard to retrain or delete models.

The workflow

What your AI does when you ask about models.

The AI acts as a bridge between your chat and your model repository.

  1. Cataloging

    The AI scans your entire library to see what is available.

    list_extraction_models

  2. Locating

    It narrows down the list to find the exact model you mentioned.

    search_models_by_name

  3. Inspecting

    The AI pulls the specific metadata and accuracy stats for that model.

    get_model_details

  4. Reporting

    It presents the metrics to you in a readable format.

    get_model_details

Try it

Copy these to start.

Use these prompts to begin interacting with your models.

Starting points

Swap out the model names and metric types with your own specific details.

Cradl AI Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_cKCc9rOwoSIb5VyjCclfT9VsRpnAzqU3YLuFHFXM/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — Cradl AI capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "cradl-ai-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_cKCc9rOwoSIb5VyjCclfT9VsRpnAzqU3YLuFHFXM/mcp"
    }
  }
}
Copy into chat04

  • List all my current extraction models.

  • Find the model named 'invoice_parser_v2' and tell me its accuracy.

  • Show me the metadata for my 'receipt_extractor' model.

  • Which models do I have available in my Cradl AI account?

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect Cradl AI once, then ask.

One connection gives your agent access to your model data. Your credentials stay encrypted and secure.

Connect Cradl AI to your AI

FAQ

How this task behaves

  • 01

    Can the AI retrain my models?

    No. The AI can only list, search, and retrieve information about your models.

  • 02

    Can I delete a model through the chat?

    No. Deleting or modifying models must be done through the Cradl AI interface.

  • 03

    Can the AI see my training data?

    No. The AI only has access to model metadata and accuracy metrics.

  • 04

    How does the AI find a specific model?

    It uses the search tool to match the name you provide against your model library.

  • 05

    What kind of metrics can I get?

    You can retrieve any accuracy metrics and metadata that are exposed by the model details tool.

  • More questions about Cradl AI? The Connector page answers them. See everything the Cradl AI Connector can do

Connect Cradl AI to Claude, Cursor, ChatGPT & more