When you ask Stevie a question, your message flows through a small but powerful system designed to produce accurate, practical, and highly personalized guitar guidance.
By combining modern asynchronous messaging with robust Retrieval-Augmented Generation (RAG), the system delivers deeply researched, long-form musical explanations straight to your inbox without interrupting your practice flow.
System Architecture Workflow
The Technical Stack
While the workflow diagram illustrates how a question travels through the system, here is a quick look at the specialized tools making it happen:
-
Chatwoot Gateway: Serves as Stevie’s communication layer, handling inbound emails and channeling them cleanly into the automated backend.
-
n8n Orchestration: The core engine that automates the entire lifecycle—managing data flow, triggering search webhooks, and structuring the final model prompts.
-
Qdrant Vector Database: Houses the curated guitar theory wiki, semantic lessons, and conversation history, enabling rapid, context-aware memory retrieval.
-
Large Language Model (LLM): Processes the optimized, context-stuffed prompt to generate Stevie’s signature structured, musical response.
Why this approach works for musicians
Traditional chat widgets require you to stay glued to a browser tab, breaking your focus. By decoupling the interface into a strictly asynchronous email loop, you can fire off a question during a practice session, keep your hands on the instrument, and return to a detailed, beautifully formatted reference guide whenever you are ready.