Description
Overview
This personal shopper automation workflow enables precise product search and store information retrieval through a combined approach of product filtering and retrieval-augmented generation (RAG). Implementing a no-code integration pipeline, it leverages chat message triggers and semantic search over embedded documents to satisfy user inquiries in real time.
Designed for e-commerce store operators and developers, this workflow addresses the challenge of interpreting customer chat inputs to deliver relevant product recommendations or store details. It utilizes a chat message webhook trigger to initiate processing, ensuring deterministic input capture and response generation.
Key Benefits
- Enables real-time chat-based product search using keyword, SKU, category, and price filters.
- Integrates a retrieval-augmented generation (RAG) system for accurate store information delivery.
- Maintains conversational context via session-specific window buffer memory for coherent interactions.
- Leverages semantic vector search with Qdrant for high-relevance document retrieval from Google Drive.
Product Overview
This personal shopper orchestration pipeline begins with a webhook listener node that receives chat messages containing a session ID and user input text. The workflow extracts these fields for structured processing. An AI-powered information extractor node analyzes the input to determine if the user intent is product-related, extracting key parameters such as keywords, SKU, category, and price ranges. Based on this analysis, the AI agent node routes the request either to a WooCommerce API query for product retrieval or to a retrieval-augmented generation system.
The RAG subsystem accesses store information documents stored in Google Drive, downloading and converting them to plain text. Text is chunked and embedded using OpenAI embeddings before indexing into a Qdrant vector database for semantic search. Queries against this vector store return relevant document snippets, which an OpenAI chat model uses to generate context-aware responses. This workflow executes synchronously from message receipt to response generation, with session-based memory nodes preserving conversational coherence. Error handling relies on default platform mechanisms without custom retry or backoff logic.
Features and Outcomes
Core Automation
This automation workflow processes chat inputs to determine user intent and extract structured search parameters for product lookup or store inquiry. The AI agent node implements decision logic to route requests conditionally within the orchestration pipeline.
- Input includes session ID and chat text, extracted by dedicated nodes.
- Decision branch based on extracted intent field “search” toggles product search or RAG response.
- Single-pass evaluation of chat input with deterministic output routing.
Integrations and Intake
The workflow integrates multiple APIs and tools to fulfill the personal shopper use case, combining e-commerce product search with document-based knowledge retrieval. Authentication is handled via API keys for OpenAI, WooCommerce, Google Drive, and Qdrant.
- WooCommerce API for product catalog queries using SKU, keyword, category, and price parameters.
- Google Drive API for accessing store documents, authenticated via OAuth2.
- Qdrant vector store for semantic indexing and search of embedded document chunks.
Outputs and Consumption
Responses are generated synchronously and returned as structured chat messages. Product search results include product details filtered by availability and extracted criteria. RAG responses provide contextually relevant text from store documentation.
- Output formats include JSON objects with product listings or textual answers.
- Responses are delivered in real time within the same session context.
- Structured keys include SKU, keyword, price range, category, and chat response text.
Workflow — End-to-End Execution
Step 1: Trigger
The workflow initiates upon receiving a chat message via a webhook listener node configured to capture session ID and chat input fields. This event-driven trigger ensures immediate processing of user interactions.
Step 2: Processing
Input data undergoes extraction and assignment of session ID and chat text fields. The information extractor node performs natural language understanding to detect product search intent and extract parameters such as keywords, SKU, category, and price limits. Basic presence checks validate required fields before proceeding.
Step 3: Analysis
The AI agent node applies conditional logic based on the extracted search flag. If true, it invokes the WooCommerce personal shopper node to perform filtered product queries. Otherwise, it queries the RAG system. The RAG system embeds the user question, queries the Qdrant vector store, and uses retrieved document content for answer generation.
Step 4: Delivery
Results, either product data or context-aware text answers, are returned synchronously to the chat interface. The workflow maintains session context for multi-turn dialogue coherence using a window buffer memory node.
Use Cases
Scenario 1
A customer inquires about available bags within a specified price range. The workflow extracts keywords and price filters, queries WooCommerce for in-stock products, and returns matching items, enabling targeted recommendations in one response cycle.
Scenario 2
A user asks for store opening hours. The workflow identifies the inquiry as informational, routes to the RAG system, retrieves relevant document excerpts from Google Drive via semantic search, and generates a precise answer referencing up-to-date store data.
Scenario 3
During a chat session, a user requests product availability by SKU. The information extractor isolates the SKU, triggering the WooCommerce product query node. The workflow returns stock status and product details while preserving session memory for follow-up questions.
How to use
To deploy this personal shopper orchestration pipeline, import the workflow into an n8n instance with configured API credentials for OpenAI, WooCommerce, Google Drive, and Qdrant. Set up the webhook trigger to receive chat messages containing session ID and user input. Ensure Google Drive contains the relevant store documents indexed in the Qdrant vector store. Once activated, the workflow runs automatically, processing each chat message to deliver structured product search results or document-based store information. Expect synchronous responses that maintain conversational context per session.
Comparison — Manual Process vs. Automation Workflow
| Attribute | Manual/Alternative | This Workflow |
|---|---|---|
| Steps required | Multiple manual lookups in product catalog and document repositories. | Automated intent detection and conditional routing with single-pass execution. |
| Consistency | Dependent on human interpretation and data retrieval accuracy. | Deterministic extraction and semantic search for reproducible responses. |
| Scalability | Limited by human availability and speed. | Handles concurrent chat sessions leveraging scalable API and vector search. |
| Maintenance | Requires manual updating of product info and store data. | Integrates automated document ingestion and indexing with minimal upkeep. |
Technical Specifications
| Environment | n8n automation platform with configured API credentials |
|---|---|
| Tools / APIs | OpenAI API, WooCommerce API, Google Drive API, Qdrant Vector Database |
| Execution Model | Synchronous request-response with session-based conversational memory |
| Input Formats | JSON payload via webhook containing sessionId and chatInput string |
| Output Formats | JSON with product listings or generated text responses |
| Data Handling | Transient processing with no persistent data storage within the workflow |
| Known Constraints | Relies on availability of external APIs and vector database service |
| Credentials | API keys and OAuth2 for OpenAI, WooCommerce, Google Drive, and Qdrant |
Implementation Requirements
- Configured API credentials for OpenAI, WooCommerce, Google Drive, and Qdrant vector store.
- Google Drive folder containing store documents properly indexed in the Qdrant collection.
- Webhook endpoint accessible for receiving chat message payloads with sessionId and chatInput fields.
Configuration & Validation
- Confirm all API credentials are valid and authorized for their respective services.
- Verify the Google Drive folder contains up-to-date store documents and is linked to Qdrant vector collection.
- Test webhook trigger by sending sample chat messages including sessionId and chatInput, ensuring proper extraction and routing.
Data Provenance
- Trigger node “When chat message received” initiates workflow on incoming chat message.
- “Information Extractor” node uses OpenAI language model to parse intent and extract parameters.
- “personal_shopper” node queries WooCommerce API; “RAG” node performs semantic search using Qdrant vector store and document embeddings.
FAQ
How is the personal shopper automation workflow triggered?
The workflow is triggered by a webhook receiving chat messages with session ID and input text, enabling event-driven processing.
Which tools or models does the orchestration pipeline use?
It uses OpenAI language models for intent extraction and response generation, WooCommerce API for product queries, Google Drive API for document retrieval, and Qdrant for vector-based semantic search.
What does the response look like for client consumption?
Responses are JSON objects containing either filtered product lists with stock status or generated text answers referencing store information documents.
Is any data persisted by the workflow?
The workflow processes data transiently; no user data or chat history is stored persistently within the workflow nodes.
How are errors handled in this integration flow?
Error handling relies on n8n platform defaults; no custom retry or backoff mechanisms are implemented within the workflow.
Conclusion
This personal shopper automation workflow delivers structured product search and document-based store information through a combined no-code integration and retrieval-augmented generation pipeline. It provides deterministic, session-aware responses by parsing user chat inputs and routing queries to appropriate APIs and semantic search services. The workflow’s reliance on external API availability and vector database uptime is a critical operational constraint. Overall, it offers a precise, maintainable solution for e-commerce conversational assistance within n8n environments.
Reviews
There are no reviews yet.