Skip to main content
This will help you get started with the Egnyte retriever. For detailed documentation of all EgnyteRetriever features and configurations head to the API reference.

​
Overview

The EgnyteRetriever class helps you search and retrieve documents from Egnyte using hybrid search capabilities that combine semantic and keyword search. This retriever is fully compliant with LangChain standards and supports both synchronous and asynchronous operations.

​
Integration details

Bring-your-own data (i.e., index and search a custom corpus of documents):
RetrieverSelf-hostCloud offeringPackage
EgnyteRetriever❌✅egnyte-langchain-connector

​
Setup

In order to use the Egnyte package, you will need:
  • An Egnyte account — If you are not a current Egnyte customer or want to test outside of your production Egnyte instance, you can use a free developer account.
  • An Egnyte app — This is configured in the developer console, and must have the appropriate scopes enabled.
  • The app must be enabled by the administrator. For free developer accounts, this is whoever signed up for the account.

​
Credentials

For these examples, we will use Bearer token authentication with an Egnyte user token. To generate a user token:
  1. Register for a developer account at https://developers.egnyte.com/member/register
  2. Generate a user token following the Public API Authentication guide
  3. Important: Use the scope Egnyte.ai when generating the token
import getpass
import os

egnyte_user_token = getpass.getpass("Enter your Egnyte User Token: ")
domain = input("Enter your Egnyte domain (e.g., company.egnyte.com): ")
If you want to get automated tracing from individual queries, you can also set your LangSmith API key by uncommenting below: 
os.environ["LANGSMITH_API_KEY"] = getpass.getpass("Enter your LangSmith API key: ")
os.environ["LANGSMITH_TRACING"] = "true"

​
Installation

This retriever lives in the egnyte-langchain-connector package: 
pip install -qU egnyte-langchain-connector

Note: you may need to restart the kernel to use updated packages.

​
Instantiation

Now we can instantiate our retriever: 
from langchain_egnyte import EgnyteRetriever

retriever = EgnyteRetriever(domain=domain, k=100)

​
Usage

query = "machine learning policies"

documents = retriever.invoke(query, egnyte_user_token=egnyte_user_token)

for doc in documents:
    print(f"Title: {doc.metadata.get('title', 'N/A')}")
    print(f"Content: {doc.page_content[:200]}...")
    print("---")

​
Advanced Search with Options

For more granular search, you can use EgnyteSearchOptions to filter results by folder path, date range, and more: 
from langchain_egnyte import EgnyteRetriever, EgnyteSearchOptions

search_options = EgnyteSearchOptions(
    limit=50,
    folderPath="/policies",
    excludeFolderPaths=["/temp", "/archive"],
    createdAfter=1640995200000,  # Unix timestamp in milliseconds (Jan 1, 2022)
    createdBefore=1672531200000  # Unix timestamp in milliseconds (Jan 1, 2023)
)

retriever = EgnyteRetriever(
    domain=domain,
    k=50,
    search_options=search_options
)

documents = retriever.invoke(
    "compliance requirements",
    egnyte_user_token=egnyte_user_token
)

​
Async usage

The retriever supports asynchronous operations: 
import asyncio

async def search_async():
    documents = await retriever.ainvoke(
        "data privacy guidelines",
        egnyte_user_token=egnyte_user_token
    )
    return documents

# Run async search
documents = asyncio.run(search_async())

​
Batch operations

You can process multiple queries in batch: 
queries = [
    "security policies",
    "employee handbook",
    "compliance guidelines"
]

# Synchronous batch
results = retriever.batch(
    queries,
    config={"configurable": {"egnyte_user_token": egnyte_user_token}}
)

# Asynchronous batch
results = await retriever.abatch(
    queries,
    config={"configurable": {"egnyte_user_token": egnyte_user_token}}
)

​
Use as an agent tool

Like other retrievers, EgnyteRetriever can be added to a LangGraph agent as a tool. 
from langchain import hub
from langchain.agents import AgentExecutor, create_openai_tools_agent
from langchain.tools.retriever import create_retriever_tool
from langchain_openai import ChatOpenAI

retriever = EgnyteRetriever(domain=domain, k=50)

egnyte_search_tool = create_retriever_tool(
    retriever,
    "egnyte_search_tool",
    "This tool searches Egnyte and retrieves documents that match the search criteria using hybrid search"
)

tools = [egnyte_search_tool]

prompt = hub.pull("hwchase17/openai-tools-agent")
llm = ChatOpenAI(temperature=0)

agent = create_openai_tools_agent(llm, tools, prompt)
agent_executor = AgentExecutor(agent=agent, tools=tools)

result = agent_executor.invoke({
    "input": "Find all documents related to data privacy policies"
})

print(result['output'])

​
API reference

For detailed documentation of all EgnyteRetriever features and configurations, visit the GitHub repository.

​
Help

If you have questions, check out the Egnyte developer documentation or reach out to the Egnyte developer community.
Edit this page on GitHub or file an issue.
Connect these docs to Claude, VSCode, and more via MCP for real-time answers.