Getting Started with Memof.ai#

Welcome to documentation that doesn't suck 🎉

What You'll Find Here#

This guide will take you from zero to AI-with-memory in minutes. Pick your path:

• 🌟 Introduction - What is Memof.ai and why should you care?
• 🧠 Core Concepts - Understand Workspaces, Bots, and Memories
• ⚡ Quick Start - Build your first app in 5 minutes
• 🏆 Best Practices - Do it right from the start
• 🔧 Troubleshooting - Fix common issues

Quick Navigation#

New to Memof.ai?#

Start with Introduction → Core Concepts → Quick Start

Ready to Build?#

Jump straight to Quick Start and pick your language!

Having Issues?#

Check Troubleshooting for solutions

Want to Master It?#

Read Best Practices for production-ready patterns

Super Quick Preview#

Can't wait? Here's the TL;DR ⚡

Memof.ai enables AI agents to maintain persistent memory across conversations. Store context, search semantically, and build intelligent applications.

Quick Install#

pip install memofai

Quick Start#

1. Get API Token#

Visit memof.ai to get your API token (format: moa_...).

2. Basic Example#

from memofai import create_moa_client

client = create_moa_client(api_token='moa_your_token')

# Create workspace and bot
workspace = client.workspaces.create({'name': 'My App'})
bot = client.bots.create({
    'name': 'Assistant',
    'workspace_id': workspace['id']
})

# Store and search memories
client.memories.create({
    'bot_id': bot['id'],
    'content': 'User prefers dark mode'
})

results = client.memories.search({
    'bot_id': bot['id'],
    'query': 'user preferences'
})

Core Concepts#

• Workspaces - Organize bots into logical groups
• Bots - AI agents with isolated memory stores
• Memories - Semantically searchable information
• Metadata - Optional structured data for filtering

Next Steps#

• Python SDK - Python integration guide
• JavaScript SDK - JavaScript/TypeScript guide
• HTTP API - Direct API reference
• MCP Server - Claude Desktop integration