# watsonx Discovery MCP for AI Agents AI Agent Connect

> watsonx Discovery lets you connect your IBM cognitive search data to your AI agent. You can query complex, unstructured datasets using natural language, check the status of your document ingestion, and monitor NLP enrichments like sentiment or entity extraction. It turns your IBM Cloud data into a searchable knowledge base for your agent to use in real-time.

## Overview
- **Category:** industry-titans
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_MOvCoAeU4n0XLtoAj5AMd7HKRadaRJgAFtxeD6Bh/ai-agent-connect
- **Tags:** cognitive-search, nlp, unstructured-data, semantic-search, text-analytics, enterprise-search

## Description

This Connector connects your enterprise cognitive search account to your AI agent to provide semantic search and NLP driven text analytics through natural conversation. Imagine you have a massive repository of enterprise documents like legal contracts, technical manuals, and internal research scattered across different folders. Finding a specific needle in that haystack usually means jumping between different dashboards, copy-pasting snippets, and hoping the search results actually make sense. This Connector changes that by letting your agent do the heavy lifting for you. Instead of manual digging, you can just ask your agent to find specific clauses or summarize findings from a specific collection. It bridges the gap between your raw data and your conversation with the AI. You get to see what is happening with your data ingestion, check which NLP models are active, and run complex queries without ever leaving your chat interface. It is about turning a mountain of data into a conversational resource. By adding this to your stack on Vinkius, you are essentially giving your agent a specialized brain for your company's unique information. You can stop worrying about the technical plumbing of the cloud console and start focusing on the actual answers you need to get your work done. You can check if your project components are healthy or see which enrichments like sentiment and entity extraction are currently active. It replaces the need to navigate complex menus with a simple chat interface, making it much easier to audit your data or find specific information quickly.

## Tools

### get_document_details
Pull up specific metadata and ingestion status for any indexed document. This helps you verify if a file was processed correctly.

### get_component_settings
View the configuration and health of your project components. Use this to troubleshoot issues with your discovery environment.

### list_discovery_collections
See every data collection within your current project. It is the quickest way to get your collection IDs.

### list_collection_documents
See every document inside a specific collection. This helps you understand the scope of what your agent can see.

### list_available_enrichments
Check which NLP models like sentiment or entities are currently active. You can use this to see how your data is being analyzed.

### query_discovery_content
Run natural language or DQL queries against your data. This is the primary way your agent fetches actual information for you.

## Prompt Examples

**Prompt:** 
```
What's in my project?
```

**Response:** 
```
I found 3 collections in your project:

| Collection Name | ID |
| :--- | :--- |
| **Legal Documents** | `col-1` |
| **Technical Support KB** | `col-2` |
| **Marketing Research** | `col-3` |

Which one would you like to query?
```

**Prompt:** 
```
Search the tech docs for battery life.
```

**Response:** 
```
I found several matches in the **Technical Support KB**. The most relevant document states:

*   **Battery Life:** Up to 12 hours of continuous use.
*   **Fast Charging:** Supports 50% charge in 30 minutes.

Would you like me to pull the full document text?
```

**Prompt:** 
```
Are my NLP models running?
```

**Response:** 
```
Your project currently has the following NLP enrichments active:

*   **Sentiment Analysis**
*   **Entity Extraction** (People, Places, Organizations)
*   **Category Classification**
*   **Keyword Extraction**

These are applied to all documents during ingestion.
```

## Capabilities

### Query collections using natural language
Perform semantic searches or DQL queries against your data using simple chat commands.

### Browse all data collections
List every collection in your project to find the right IDs for your agent to target.

### Get document metadata
Retrieve specific details and ingestion statuses for any indexed document in your set.

### Monitor NLP enrichments
Check which models like sentiment or entities are currently active on your data.

### Verify component health
Check the configuration and status of your project components to ensure everything is running.

### Surface hidden patterns
Extract relevant information from large unstructured datasets through conversational prompts.

## Use Cases

### Legal document review
A legal team needs to find all termination clauses in a 'Contracts' collection. They ask their agent to search for specific language, and the agent uses `query_discovery_content` to find and summarize the relevant sections.

### Knowledge base summarization
A support lead wants a summary of common issues in a 'Knowledge Base' collection. The agent scans the collection and provides a bulleted list of frequent problems found in the indexed documents.

### Ingestion health auditing
A data scientist needs to check if a new batch of documents was ingested correctly. They ask the agent to check the status of specific IDs using `get_document_details` to verify the project health.

### NLP enrichment monitoring
A product manager wants to see what NLP enrichments are active. They ask the agent to list the models, and it uses `list_available_enrichments` to show the sentiment and entity tags being applied.

## Benefits

- Stop manual dashboard hopping by using `query_discovery_content` to get answers directly in your chat. This keeps you in your workflow while your agent handles the complex query logic.
- Verify your data is ready for use by checking ingestion status with `get_document_details` instantly. You can confirm files are processed without manual checks.
- Monitor your NLP pipeline health with `list_available_enrichments` to ensure sentiment and entities are being captured. This ensures your data analysis is always accurate.
- Get a bird's-eye view of your project structure using `list_discovery_collections` to manage multiple data sources. It makes it easy to see exactly what your agent can access.
- Troubleshoot project configurations quickly with `get_component_settings` to keep your discovery environment running smoothly. You can identify issues before they impact your results.

## How It Works

The bottom line is that you get a conversational interface for your entire cognitive search environment.

1. Connect your watsonx URL, API Key, and Project ID.
2. Link the Connector to your preferred AI client.
3. Ask your agent to query, list, or analyze your collections.

## Frequently Asked Questions

**How does watsonx Discovery help with large data?**
It connects your massive data repositories to your AI agent, allowing you to query them using natural language instead of complex manual searches.

**Can I use watsonx Discovery for sentiment analysis?**
Yes, the Connector allows your agent to check which NLP enrichments, like sentiment analysis, are active and being applied to your data.

**How do I check if my documents were ingested?**
You can ask your agent to check the status of specific documents, and it will retrieve the ingestion status and metadata for you.

**Can I search using plain English?**
Yes, this Connector enables your agent to perform both natural language queries and DQL queries against your discovery collections.

**How do I see my project health?**
Your agent can verify project-level configurations and health settings for all your project components through simple chat commands.

**Is watsonx Discovery good for enterprise search?**
It is specifically designed for enterprise-grade cognitive search, making it ideal for searching unstructured data like contracts and manuals.

**Can I query my data collections using natural language?**
Yes. The `query_discovery_content` tool allows your AI agent to perform natural language queries against your watsonx Discovery collections, returning highly relevant results based on IBM's cognitive search engine.

**How do I see what NLP models are being applied to my documents?**
Use the `list_available_enrichments` tool to see all NLP enrichments (like Sentiment, Entity Extraction, or Category Classification) configured for your project and applied during the ingestion pipeline.

**Can I monitor the ingestion status of a specific document?**
Absolutely. Using the `get_document_details` tool, you can check the ingestion status and technical metadata for any specific document ID, ensuring your data is correctly indexed and searchable.