This project provides a self-hostable CRM template built for the WhatsApp Business API, targeting businesses and power users seeking full data ownership and customization without SaaS lock-in. It offers a comprehensive suite of features including a shared inbox for multiple agents, contact management, sales pipelines, broadcast capabilities, no-code automations, and an AI-powered reply assistant, enabling users to deploy and tailor a robust customer relationship management system.

How It Works

The CRM utilizes a modern, "boring stack" for ease of adoption: Next.js 16 (App Router) for the frontend, Supabase (Postgres, Auth, Storage) for the backend, and the official Meta Cloud API for WhatsApp integration. Its core advantage lies in its template nature, allowing users to fork, brand, and deploy their own instance with full control over code and data. Security is emphasized through robust primitives like AES-256-GCM token encryption, Row Level Security (RLS) on all database tables, and HMAC-verified webhooks. The AI assistant integrates with user-provided OpenAI or Anthropic keys, ensuring data privacy and avoiding per-seat fees.

Quick Start & Requirements

To get started, fork the repository on GitHub, clone it locally, run npm install, copy and populate the .env.local.example file with Supabase and Meta credentials, and then execute npm run dev to run the application locally on http://localhost:3000. Key prerequisites include Node.js, a Supabase project, Meta Cloud API credentials, and optionally an OpenAI or Anthropic API key for AI features. Deployment is streamlined via Hostinger's one-click Git deploy, though the application is compatible with any Node.js environment like Vercel or a custom VPS. Official documentation and deployment guides are available at wacrm.tech.

Highlighted Details

• Supports multiple agents managing a single WhatsApp number with per-conversation assignment and notes.
• Features a visual no-code automation builder triggered by messages, contacts, keywords, or schedules.
• AI reply assistant can draft responses using user-provided OpenAI/Anthropic keys and answer queries from a knowledge base using hybrid retrieval (Postgres FTS or pgvector).
• Offers real-time dashboards for key performance metrics like response times, volume, and pipeline value.
• Includes team account management with role-based access control and a public REST API for custom integrations.

Maintenance & Community

This project is designed as a template forking and customization, rather than a collaborative product. The primary expected workflow is fork → customize → deploy. While bug reports and security issues are welcomed, feature pull requests are generally expected to be implemented within user forks. No specific community channels like Discord or Slack are listed.

Licensing & Compatibility

The project is released under the MIT License, permitting free use, modification, and distribution. This license ensures broad compatibility for commercial use and integration into closed-source projects without restrictive copyleft obligations. The application runs on any platform supporting Node.js.

Limitations & Caveats

As a template, the project requires users to fork and configure it for their specific needs; it is not an out-of-the-box SaaS product. Deployment relies on external services such as Supabase, Meta Cloud API, and optionally OpenAI/Anthropic, which incur their own costs and management requirements. While Hostinger is the recommended deployment path for ease of use, users are responsible for managing infrastructure if deploying elsewhere.