Integration

Catalogian + LlamaIndex

LlamaIndex is a data framework for connecting LLMs to external data sources. Catalogian's MCP server gives LlamaIndex agents live access to product catalog data — schemas, filtered rows, change history, and downloads — without building custom data connectors or loaders.

Instead of indexing static product CSVs, connect LlamaIndex to Catalogian's MCP endpoint for always-current data. Your agents can query the latest snapshot, check what changed since yesterday, or filter products by any field — all in real time. The data is always fresh because it comes directly from your ingested feeds.

LlamaIndex's tool-use capabilities let agents call Catalogian's 14 MCP tools as part of their reasoning. Combine catalog data with other LlamaIndex data sources — knowledge bases, documents, databases — for multi-source RAG pipelines that include live product data.

Start monitoring free Read the docs

How it works

Connect to Catalogian's MCP server

Use LlamaIndex's MCP client to connect to POST /v1/mcp with your API key. All 14 tools are auto-discovered and available to your agent.

Build agents or query engines

Create LlamaIndex agents that use Catalogian tools alongside other data sources. Or build query engines that combine product data with your existing knowledge base.

Query live product data

Your LlamaIndex application gets real-time access to product catalogs, change deltas, and historical snapshots — always current, always structured.

LlamaIndex agent with Catalogian tools

from llama_index.tools.mcp import BasicMCPClient, McpToolSpec
from llama_index.agent.openai import OpenAIAgent

mcp_client = BasicMCPClient(
    "https://catalogian.com/v1/mcp",
    headers={"x-api-key": "cat_sk_live_..."}
)
mcp_tools = McpToolSpec(client=mcp_client)
tools = mcp_tools.to_tool_list()

agent = OpenAIAgent.from_tools(tools, verbose=True)
response = agent.chat(
    "Show me all products that went out of stock today "
    "in the main-catalog source"
)
print(response)

Frequently asked questions

Do I need to index product data with LlamaIndex?

Not for live queries. Catalogian's MCP tools provide direct access to current product data. Use LlamaIndex indexing for historical analysis or combining product data with other document sources.

Can I use Catalogian data in a RAG pipeline?

Yes. Catalogian's MCP tools can be part of a LlamaIndex agent that also queries a vector index. This lets you combine structured product data with unstructured knowledge base content.

What models work with the LlamaIndex + Catalogian integration?

Any model that supports tool use. This includes OpenAI GPT-4o, Anthropic Claude, and other models that LlamaIndex's agent framework supports for function calling.

Start monitoring your product feeds

Free plan includes 1 source and 50,000 SKUs. No credit card required.

Start monitoring free