​
Welcome to duohub

duohub is the foundational memory layer for voice AI applications, providing ultra-low latency knowledge graph capabilities through our API and SDK. Our service enables efficient memory management and retrieval for AI applications with global distribution.

​
Core Technology

duohub offers two primary approaches to memory management:
  • Knowledge Graph Memory: Optimized for precise memory retrieval with relationship context
  • Vector Memory: Designed for quick retrieval across large memory sets
Our globally distributed infrastructure ensures minimal latency regardless of your users’ location.

​
Data Ingestion Capabilities

duohub supports multiple data ingestion methods:

Documents

Support for PDF, Word, and Text documents

Web Content

Single page or entire website ingestion

Video

Multi-speaker video content processing

Audio

Multi-speaker audio processing

​
Integration Options

​
Python SDK

The fastest way to get started is with our Python SDK: 
from duohub import Duohub

client = Duohub(api_key='YOUR_API_KEY')
response = client.query(
    query='What were Leonhard Euler\'s contributions to graph theory?',
    memoryID='your_memory_id',
    facts=True,
    assisted=True
)

​
REST API

For other languages, use our REST API with comprehensive OpenAPI specification.

Quick Start Guide

Set up your first knowledge graph in minutes

API Reference

Explore our API documentation

​
Key Features

Knowledge Graphs

High-performance graph architecture for precise memory retrieval

Memory Retrieval

Assisted and non-assisted retrieval with supporting facts

Global Distribution

Ultra-low latency serving through distributed infrastructure

Use Cases

Implementation examples and best practices