WhatsApp AI Assistant Bot (MVP)

An AI assistant that runs inside WhatsApp. It can reply to messages, speak with voice notes, transcribe audio, generate and analyze images, summarize group chats, moderate users, and remember personal facts.

Key Features

• 
  /ai <message>
  – Smart text replies.
  


• 
  /voice <message>
  – Text-to-Speech voice note.
  


• 
  /listen
  – Speech-to-Text for voice notes.
  


• 
  /img <prompt>
  – AI image generation.
  


• 
  /describe
  – Describe an image.
  


• 
  /ocr
  – Extract text from images.
  


• 
  /mode <persona>
  – Switch persona (formal, funny, developer, romantic, teacher, sheikh).
  


• 
  /remember <info>
  /
  /recall
  – Save and view user facts.
  


• 
  /summary
  /
  /autosummary on|off
  – Chat summaries.
  


• 
  /mute @user <minutes>
  /
  /unmute @user
  – Moderation (admin only).
  

AI Capabilities

• Fluent, context-aware conversation.


• User memory per contact.


• High-accuracy voice transcription (Whisper).


• Natural voice replies (TTS).


• Vision, OCR, and image understanding.


• Group chat summarization.

Tech Stack

• 
  Node.js (ESM),
  whatsapp-web.js
  


• Groq API (LLaMA 3.3 70B) for language


• OpenAI Whisper (STT), Google TTS


• Pollinations AI (image generation)


• SQLite (persistent memory)


• Socket.IO + simple Web UI (dashboard)


• Regex filters + mute state machine (moderation)

Running Examples

/ai explain quantum computing simply

/voice write a bedtime story about the moon

/img futuristic cyberpunk mosque at sunrise

/remember my favorite color is blue

/summary

Project Structure

/core        → AI, STT, TTS, Vision

/whatsapp    → Session, Commands, Moderation

/db          → Memory, personas, logs, mute states

/dashboard   → Live monitoring UI

server.js    → Entry

bot.db       → SQLite

Safety

• Profanity filter (AR/EN), anti-spam, group-safe mode.

Scalable as a SaaS system

Users can sign up, scan their QR, and get an isolated assistant. Multi-instance support included.

/ai اشرح الحوسبة الكمومية ببساطة

/voice اكتب قصة ما قبل النوم عن القمر

/img مسجد سيبر بانك مستقبلي عند الشروق

/remember لوني المفضل أزرق

/summary

/core        → الذكاء، الصوت، الرؤية

/whatsapp    → الجلسة، الأوامر، الإشراف

/db          → الذاكرة، الشخصيات، السجلات، حالات الكتم

/dashboard   → واجهة المتابعة الحية

server.js    → نقطة الدخول

bot.db       → SQLite