Handle customer support queries with cache-first RAG using Redis, LangCache and OpenAI

An end-to-end Retrieval-Augmented Generation (RAG) customer support workflow for n8n, using a cache-first strategy (LangCache) combined with a Redis vector store powered by OpenAI embeddings. This template is designed for fast, accurate, and cost-efficient customer support chatbots, internal help desks, and knowledge-base assistants.

---

Overview

This workflow implements a production-ready RAG architecture optimized for customer support use cases. Incoming chat messages are processed through a structured pipeline that prioritizes cached answers, falls back to semantic vector search when needed, and validates response quality before returning a final answer.

The workflow supports: - Multi-question user inputs - Intelligent query decomposition - Cache reuse to reduce latency and cost - High-precision retrieval from a Redis vector database - Quality evaluation and controlled retries - Final answer synthesis into a single, coherent response

---

Key Features

- Chat-based RAG pipeline using n8n’s Chat Trigger - Query decomposition for multi-topic questions - LangCache integration (search + save) - Redis Vector Store for semantic retrieval - OpenAI embeddings and chat models - Quality scoring with retry logic - Session memory buffers for contextual continuity - Fallback-safe behavior (no hallucinations)