Makinari API
Welcome to the Makinari API documentation. This platform lets you deploy autonomous AI agents across multiple channels.
What is Makinari?
Makinari gives you a fleet of specialized AI agents that work autonomously toward your business goals. Instead of building one-off automations, you define high-level objectives and let agents plan, execute, and iterate on their own.
Agents operate across channels (WhatsApp, email, website chat, social), use tools exposed via the MCP server, and can be orchestrated programmatically through the REST API.
How it fits together
- Channels are the surfaces where your agents talk to customers. Each channel has its own setup and configuration.
- Agents are the workers. Each one is specialized for a role: sales, support, copywriting, analytics, and more.
- AI Goals (Requirements) are the objectives you assign. Agents consume a token budget to pursue them autonomously.
- REST API lets your own systems read and write data: visitors, contacts, robot instances, and more.
- MCP Server exposes Makinari tools directly to AI models like Claude or GPT, enabling model-driven workflows.
Where to start
- First Steps - Everything you need to get started.
- Channels - Connect with your audience where they are.
- Agents - Deploy specialized workforce agents.
- REST API - Deep integration with your systems.
- MCP Server - Extend capabilities with the Model Context Protocol.
- Open Source Repos - Explore our open source ecosystem.
New here? Start with First Steps to create your account, configure your project, and get your API key.
Last updated on