I was recently informed that Meta is rolling out Business AI for Messenger and WhatsApp.
So far, I’ve only been able to get it working on Messenger. I still haven’t successfully integrated it with WhatsApp Business Accounts (WABA).
Has anyone here already integrated Business AI on WhatsApp successfully? Any tips, requirements, or setup guidance would be greatly appreciated.
Embedded Signup is very convenient and easy to use, but it also means you usually have to operate through a Tech Provider or BSP ecosystem.
I’m currently exploring using the official WhatsApp Cloud API without Embedded Signup because I want full control over:
• Meta app configuration
• Access tokens and system users
• Webhooks and infrastructure
• Self-hosted deployments
Has anyone here successfully set this up manually in production? Would love to hear your experience or any pitfalls to avoid.
Hi all 👋
I’m one of the builders behind ChatbotX and would love to get feedback, ideas, and hopefully some contributors from the community.
https://github.com/ChatbotXIO/ChatbotX
ChatbotX is an open-source omnichannel chatbot platform similar to ManyChat, Wati, Chatfuel, and Respond io, but designed to be self-hostable, extensible, and developer-friendly.
Features:
• WhatsApp, Facebook, Instagram, Telegram, Zalo, Email, and Webchat integrations
• Visual Flow Builder with 15+ node types
• AI Agents powered by OpenAI, DeepSeek, and Gemini
• Live Chat Inbox with human takeover and conversation assignment
• Contact CRM with tags, custom fields, and segmentation
• Broadcasting and automated sequences
• Team management and role-based permissions
• Rich messaging with buttons, quick replies, catalogs, and carousel cards
• Instagram Comment-to-DM automation
• A/B Testing and workflow triggers
• Webhooks & HTTP integrations
• Growth tools and analytics
• APIs, CLI, and MCP support for advanced agent workflows
Tech stack:
• Node.js 24 + TypeScript 5
• pnpm workspaces + Turborepo
• Next.js 16 + React 19
• PartyKit / PartySocket for realtime messaging
• PostgreSQL + pgvector + Drizzle ORM
• Redis + BullMQ for queues
• ClickHouse for analytics
• RustFS / S3-compatible storage
• Docker Compose for local infrastructure
License: AGPLv3
The project is built with clean architecture and skills/prompts to make customization, contribution, and vibe coding easier. Would love to hear your thoughts, criticism, or feature ideas!
