FOCUSCurrently working on ScheduleSomething and LayerdEvents algorithmsLearn More
FOCUSCurrently working on ScheduleSomething and LayerdEvents algorithmsLearn More
FOCUSCurrently working on ScheduleSomething and LayerdEvents algorithmsLearn More
FOCUSCurrently working on ScheduleSomething and LayerdEvents algorithmsLearn More
MetapilotGetting Started
System Overview
DocsmetapilotGetting StartedSystem Overview
GitHub Live Sync

System Overview

Live technical documentation fetched from GitHub repository omghante/metapilot/docs/getting-started/overview.md

MetaPilot Overview

MetaPilot is an enterprise-grade, Multi-Tenant SaaS platform engineered for WhatsApp Business Automation, AI-Powered Customer Engagement, High-Throughput Broadcast Campaigns, and Realtime Omnichannel Messaging.
Designed with a Modular Monolith backend architecture (Django 5 ASGI + Daphne) and a modern Next.js 16 (Turbopack) frontend, MetaPilot enables marketing agencies and enterprise client accounts to manage Meta WhatsApp Business API channels at scale.

Core Capabilities

1. Multi-Tenant Account Hierarchy

  • Super Admin: Platform owners managing agencies, global system configuration, audit logs, and metrics.
  • Agency Admin: Marketing agencies managing client accounts under their agency subscription.
  • Tenant Admin: Client organization admins managing team members, WhatsApp Business Accounts (WBA), and broadcast templates.
  • Tenant User: Operators monitoring chats, replying to customer conversations, and scheduling broadcasts.

2. Realtime Chat Inbox

  • Low-latency WebSockets powered by Django Channels and Redis Channel Layer.
  • Instant delivery of inbound WhatsApp messages and outgoing responses.
  • Unread counter tracking and live agent assignment.

3. High-Performance Broadcast Campaign Engine

  • Asynchronous broadcast scheduling powered by Celery Beat and Redis.
  • Token Bucket Rate Limiting (up to 50 msgs/sec per tenant) ensuring compliance with Meta WhatsApp Graph API limits.
  • Automated retry mechanisms with exponential backoff for failed deliveries.

4. AI-Powered Auto-Responders & Platform Assistant

  • Dual AI engines integrated via OpenRouter:
    • WhatsApp AI Auto-Responder: Handles inbound customer inquiries using tenant-specific knowledge bases and vision models (GPT-4o / StepFun 3.5 Flash).
    • Platform AI Assistant: Answers operator queries regarding platform features, analytics, and campaign setup.

5. Multi-Tenant Cryptographic Isolation

  • Cryptographic secret isolation per tenant using Fernet (AES-256 in CBC mode).
  • Secure storage of Meta API access tokens, phone number IDs, and webhook verification secrets.

Technology Stack

  • Frontend: Next.js 16 (React 19, Turbopack, Vanilla CSS, TypeScript).
  • Backend API: Python 3.14 / 3.11, Django 5.0, Django REST Framework, SimpleJWT.
  • ASGI & WebSockets: Daphne 4.0, Django Channels 4.0, Redis 7.0+.
  • Background Worker & Scheduler: Celery 5.3, Django Celery Beat.
  • Database: SQLite (Local Dev) / PostgreSQL 15+ (Production).
  • Cryptography: Fernet AES-256 (
    cryptography
    ).