Conversational WhatsApp Assistant with Gemini AI and Pinecone Memory

Go to Workflow
0 views
Built by Adrian Adrian
Created on July 19, 2025

Description

šŸ“‹ Description

This template creates an intelligent AI assistant for WhatsApp that can:
Respond naturally** to messages using Google Gemini AI
Remember previous conversations** for each user
Access a knowledge base** for answering frequently asked questions
Automatically save** all conversations for long-term memory

šŸ› ļø Requirements

1. WAMM.pro Account (FREE tier available)
What is WAMM.pro?** - A platform that enables WhatsApp automation using proprietary API technology
Free tier:** 50 messages/month
PRO tier:** Unlimited messages + advanced features
Link:** wamm.pro

2. Pinecone Account (for AI memory)
For storing conversations and knowledge base
Free tier available

3. Google AI Account (for Gemini)
For the conversational AI model

4. OpenAI Account (for embeddings)
For generating memory vectors

šŸš€ Step-by-step Setup

Step 1: WAMM.pro Configuration

Create account at wamm.pro
Account Manager → Add WhatsApp profile
Scan QR code with your WhatsApp
Note down: Instance ID and Access Token

Step 2: Webhook Configuration

In WAMM.pro: Integrations → Webhooks → Messages Webhooks
Add Webhook with the n8n URL
Required configuration:
From others: āœ… Relevant + āœ… Without media + āœ… Exclude no text
To others: āœ… Relevant + āœ… Without media + āœ… Exclude no text
To myself: āœ… None (to avoid responding to own messages)

Step 3: Pinecone Configuration

Create 2 indexes:
historywa - for conversation memory
knowledge - for knowledge base
Index settings:
Dimensions: 3072
Metric: cosine
Embedding model: text-embedding-3-large

Step 4: n8n Configuration

Configure credentials:
WAMM: Instance ID + Access Token
Pinecone: API Key
Google Gemini: API Key
OpenAI: API Key for embeddings

šŸ”§ How it Works

Workflow Flow:

šŸ“± WhatsApp Message
↓ (webhook)
šŸŽÆ AI Agent (Gemini)
↓ (uses tools)
šŸ“š Memory Tool + Knowledge Tool
↓ (response generated)
šŸ“¤ WAMM Send Message
↓ (saves)
šŸ’¾ Pinecone Memory Storage

Available AI Tools:

Memory Tool - Searches previous conversations with the user
Knowledge Tool - Searches the general knowledge base

Special Features:

Natural conversations** - AI doesn't mention "searching history"
Persistent context** - Remembers names, preferences, previous conversations
User language detection** - Automatically responds in user's language
Organized memory** - Each user has their own memory space

šŸ“Š Benefits

āœ… Zero maintenance - Runs automatically
āœ… Scalable - Supports multiple users simultaneously
āœ… Intelligent memory - Uses similarity search for relevant context
āœ… Extensible - Easy to add new features
āœ… Cost-effective - Free tiers available for all services

šŸŽÆ Use Cases

Automated customer support** with memory
Personal assistant** for WhatsApp
Business chatbot** with specific knowledge
Conversation automation** with persistent context

šŸ”’ Security

Data** stored in Pinecone as vector embeddings
No plain text** message storage
Each user** has separate memory space
API keys** secured in n8n credentials

šŸ“ˆ Possible Extensions

CRM** integrations
Scheduling** and reminders
Advanced multi-language** support
Analytics** and conversation reports
Custom knowledge bases** per user

šŸ’” Tip: For optimal results, populate the knowledge base with frequently asked questions specific to your business!

Nodes Used (7)

AI Agent
@n8n/n8n-nodes-langchain.agent
Code
n8n-nodes-base.code
Default Data Loader
@n8n/n8n-nodes-langchain.documentDefaultDataLoader
Embeddings OpenAI
@n8n/n8n-nodes-langchain.embeddingsOpenAi
Google Gemini Chat Model
@n8n/n8n-nodes-langchain.lmChatGoogleGemini
Pinecone Vector Store
@n8n/n8n-nodes-langchain.vectorStorePinecone
Recursive Character Text Splitter
@n8n/n8n-nodes-langchain.textSplitterRecursiveCharacterTextSplitter