How Makinari Fits Together

Makinari is a platform that deploys autonomous AI agents to interact with your customers and execute business tasks. Instead of static, rule-based automations, Makinari operates on an objective-based architecture.

The Architecture

1. The Site

Everything in Makinari happens within a Site. A Site represents a business, project, or workspace. It holds your brand context, leads, campaigns, and configuration.

2. Requirements (AI Goals)

Instead of building workflows block-by-block, you define Requirements. A Requirement is a high-level goal (e.g., “Research these leads and draft personalized outreach emails”).

3. The CMO Agent

When a Requirement is activated, the CMO Agent acts as the orchestrator. It reads the goal, analyzes the budget, creates a plan, and delegates specific tasks (Commands) to the specialized workers.

4. Specialized Agents

Makinari has a fleet of role-specific agents:

Sales: Generates and qualifies leads.
Copywriter: Drafts and edits content.
Data Analyst: Enriches lead data and builds segments.
Support: Answers customer queries based on context.

5. Tools and Channels

To execute their tasks, agents use Tools (like searching the web, querying the CRM, or sending an email). They communicate with the outside world through Channels (Website chat, WhatsApp, Email).

How you interact with it

You can interact with this system in three ways:

The Dashboard: Use the visual UI at app.makinari.com to configure Sites, Channels, and Requirements.
The REST API: Trigger workflows, sync leads, and pull reports programmatically.
The MCP Server: Connect your local AI tools (like Cursor) directly to the Agent Tools.