Google Play Review Summarizer with GPT, Pinecone & Slack - n8n Workflow

Markdown formatted list:
Product Managers who need automated, high-level summaries of user sentiment.
Community Managers monitoring app reviews and user feedback trends.
Data Analysts seeking ready-to-use n8n templates for RAG-based analysis.
Developers needing a robust n8n workflow solution for app monitoring.

• Anyone looking to leverage the power of n8n, OpenAI, and Pinecone together.

Keeping up with app reviews can be time-consuming, but ignoring them leads to missed opportunities. This sophisticated n8n workflow solves this problem by automating the entire analysis process. It utilizes a powerful Retrieval Augmented Generation (RAG) architecture: daily reviews are fetched and stored as vectors in Pinecone, creating a specialized knowledge base. A separate, weekly n8n trigger then activates an OpenAI Agent (using gpt-4-mini) to analyze this data. This n8n workflow generates a comprehensive summary, including star rating breakdowns and sentiment analysis (positive vs. negative feedback), before posting the concise results directly to a designated Slack channel. This robust use of the n8n node ecosystem ensures you never miss critical user feedback.

This n8n workflow operates via two main paths, managed by two distinct schedule triggers:

Path 1: Daily Data Collection and Storage (RAG Preparation)

1. Daily Trigger: The n8n trigger fires daily at 10 am, initiating the data collection process.


2. App List Processing: The workflow retrieves a list of targeted Google Play bundle IDs and loops over them (Loop Over Items).


3. Review Fetching: An HTTP Request n8n node connects to the Google Publisher API (requiring a Google Service Account credential) to pull all available reviews for the current app ID.


4. Filtering: A Filter n8n node isolates only yesterday's reviews, ensuring the vector database remains up-to-date.


5. Vectorization: Reviews are prepared by the Default Data Loader (including metadata like star rating and date), converted into vector embeddings via the Embeddings OpenAI n8n node, and then inserted into the Pinecone Vector Store. Critically, this n8n node uses dynamic namespaces per app, and the namespace is automatically cleared every Saturday to maintain data freshness.

Path 2: Weekly Summary Generation and Notification

1. Weekly Trigger: A second n8n trigger fires every Friday at 11 am, initiating the summary process.


2. App List Processing: The workflow loads app names and bundle IDs and loops over them (Loop Over Items1).


3. Data Retrieval: The Pinecone Vector Store1 n8n node acts as a tool, retrieving the relevant review vectors (up to 500 documents) for the current app using Embeddings OpenAI1.


4. AI Analysis: The AI Agent - Summary creator receives the retrieved documents. Coupled with the OpenAI Chat Model1 (configured to use GPT-4-mini), it executes a detailed system prompt to create a structured summary (positive/negative sentiment, star rating analysis).


5. Notification: Finally, the Send to Slack channel n8n node posts the AI-generated summary, providing immediate, actionable insights to your team.

This complex n8n workflow requires several premium credentials to function correctly. To import and use this specific n8n templates guide:

1. Import: Copy the provided JSON into your n8n instance using the 'New' -> 'Import from JSON' option.


2. Google Service Account: Configure a Google API credential for the HTTP Request n8n node to access the Google Play Store reviews API.


3. OpenAI: Set up an OpenAI credential for both the Embeddings OpenAI n8n node and the OpenAI Chat Model1 n8n node. This will handle both vectorization and summary generation.


4. Pinecone: Configure a Pinecone API credential and ensure you have an index named n8n-store-reviews where the vector data will reside.


5. Slack: Configure a Slack API credential and update the Send to Slack channel n8n node with the correct Channel ID where you wish the summaries to be posted.


6. Configuration Nodes: Modify the Set the bundle ids and Set app details n8n nodes to include the actual bundle IDs and corresponding names for the apps you are monitoring.


7. Activate: Save and activate the n8n workflow to enable the scheduled daily data collection and weekly reporting via the two distinct n8n trigger nodes.

This n8n workflow leverages specialized AI and cloud service n8n node types:

Daily Trigger / Weekly trigger (n8n trigger): Defines the two automated start times. The Daily trigger fetches data, and the Weekly trigger initiates the summary process.
HTTP Request: Function: Fetches raw review data from the Google Play Store API using a Google Service Account. Key Configuration: Uses pagination to fetch up to 10 pages of reviews.
Filter: Function: Ensures only reviews from the previous day are processed and stored. Key Configuration: Date-time expression comparing lastModified.seconds with {{$today.minus({days: 1})}}.
Default Data Loader: Function: Structures the raw review data into a format suitable for vectorization, including adding crucial metadata like starrating, date, and appversion.
Embeddings OpenAI: Function: Converts the structured review text into high-dimensional vector embeddings.
Pinecone Vector Store (Insert Mode): Function: Stores the newly generated vectors in Pinecone. Key Configuration: Uses the app bundle ID to create a unique namespace (googleplaystorereviews{{ $json.apps }}). Includes logic to clear the namespace weekly (clearNamespace).
Pinecone Vector Store (Retrieve Mode): Function: Acts as a Retrieval Tool for the AI Agent, querying the stored vectors (up to 500 results) needed for summarization.
OpenAI Chat Model1: Function: Specifies the LLM (GPT-4-mini) used by the AI Agent for processing the retrieved information.
AI Agent - Summary creator: Function: Orchestrates the RAG process. It receives the Pinecone tool output and uses a defined system message and prompt to generate a split summary (positive/negative) with star rating metrics. This n8n node is central to the analysis part of the n8n workflow.
Send to Slack channel: Function: Transmits the final AI-generated summary message to the selected Slack channel.