RAG with LlamaIndex - Integrations

Add Desearch retrieval to a LlamaIndex agent: register Desearch's search tool as a retriever, then set up an OpenAI Agent that uses it.

Quick Start

bash

pip install llama-index llama-index-core llama-index-desearch

Set the DESEARCH_API_KEY environment variable first. Get a key at Desearch.

python

from llama_index_desearch.tools import DesearchToolSpec
import os

desearch_tool = DesearchToolSpec(
    api_key=os.environ["DESEARCH_API_KEY"],
)

This example passes only the ai_search_tool method to the agent, using the .to_tool_list method:

python

search_and_retrieves_tool = desearch_tool.to_tool_list(
    spec_functions=["ai_search_tool"]
)

python

from llama_index.agent.openai import OpenAIAgent

agent = OpenAIAgent.from_tools(
    search_and_retrieves_tool,
    verbose=True,
)

Use the chat method to interact with the agent:

python

agent.chat(
    "Can you summarize the news from the last month related to the US stock market?"
)

ai_search_tool: AI-powered search across multiple sources, including web pages, research papers, and social media.
twitter_search_tool: retrieves links and tweets matching a search query, without AI models.
web_search_tool: plain web search, like a normal search engine.