Automated Google Review Sentiment Analysis and Retrieval Augmented Generation (RAG) using Pinecone - n8n Workflow

• Digital marketing agencies needing automated reputation management.


• Business owners who monitor customer feedback daily.


• n8n users looking for advanced RAG implementation examples.


• Developers needing robust n8n templates integrating LangChain, OpenAI, and Pinecone.

This comprehensive n8n workflow provides a dual-purpose automation solution for handling customer feedback. First, it establishes a daily monitoring pipeline to scrape new Google Maps reviews, use an advanced AI model (GPT-4o) for sentiment analysis, and log the results into Google Sheets. Second, it implements a full Retrieval Augmented Generation (RAG) system. Every review analyzed by this n8n node sequence is vectorized using OpenAI embeddings and ingested into a Pinecone vector database. This database powers a sophisticated AI Data Analyst accessible via a Telegram channel. Users can query the agent about historical reviews, trends, or specific topics, making this one of the most powerful n8n templates for actionable insight generation. The workflow also includes a scheduled n8n trigger for weekly database cleanup, ensuring the vector index remains efficient.

The n8n workflow operates via three main logical paths:

1. Daily Review Monitoring (Scheduled Trigger):

1. Telegram RAG Interaction (Interactive Trigger):

1. Weekly Maintenance (Scheduled Trigger):

To deploy this n8n workflow, follow these steps:

1. Import: Copy the provided JSON and paste it into your n8n instance using the 'New' -> 'Import from JSON' function.


2. Credentials: Set up credentials for:

1. Configuration Nodes: Update the ⚠️ CONFIGURATION Set n8n node series (1, 2, and the primary one) to include your specific parameters: Chat ID, Google Maps review page link, and the desired Pinecone namespace name.


2. Activation: Ensure all triggers (Trigger at midnight, Telegram Trigger, etc.) are activated and the entire n8n workflow is marked 'Active' for automatic execution.

Trigger at midnight / When clicking "Execute" (Schedule / Manual n8n trigger): Initiates the daily data retrieval and processing routine.
Telegram Trigger (n8n trigger): Activates the Retrieval Augmented Generation (RAG) query pipeline upon receiving a message, making the n8n workflow interactive.
Start Google Maps Scraper / Get Dataset Results (HTTP Request n8n node): Handles communication with the external scraping API to fetch the latest Google reviews.
Log data (Google Sheets n8n node): Archives the raw scraped review data into a spreadsheet for long-term record-keeping.
Loop Over Items (Split In Batches n8n node): Critical flow control to iterate through potentially large sets of reviews one by one or in small batches for controlled AI analysis and ingestion.
AI Review Analyst (LangChain Agent n8n node): Performs highly structured sentiment analysis and review summarization using a configured language model and output parser, a key component of this n8n workflow.
Embeddings OpenAI / Embeddings OpenAI1 (LangChain Node): Generates vector embeddings of the review content and user queries, essential for the semantic search capabilities of the n8n templates.
Pinecone Ingest / Pinecone Search Tool (LangChain Vector Store n8n node): Manages the vector database. Pinecone Ingest stores the vector data, while the Pinecone Search Tool enables the AI agent to retrieve relevant context during a RAG query.
AI Data Analyst (LangChain Agent n8n node): The RAG engine that utilizes the Pinecone knowledge base to answer user questions sent via Telegram, showcasing advanced n8n templates functionality.
Send Alert / Send a text message (Telegram n8n node): Provides real-time notifications for critical review findings and sends responses generated by the AI Data Analyst.