AddictionTube RAG FAQ is an AI-powered semantic FAQ system designed by James Murray for people in recovery, families looking for help, and professionals who need fast, accurate information about treatment and support options.

Instead of scrolling through static FAQ pages, users ask questions in their own words. The system retrieves relevant answers from curated addiction and recovery content using vector embeddings and Retrieval-Augmented Generation (RAG), then synthesizes a clear, grounded response.

How It Works

• Semantic Retrieval: User questions are embedded into vector space and matched against a large library of recovery FAQs and articles.
• Answer Synthesis: The system pulls the top matches and generates a consolidated answer, with source passages available for context.
• Domain Awareness: The content is tightly focused on addiction, treatment options, recovery support, and practical logistics.

The focus is on clarity and signal over noise—especially important when someone is in crisis, exhausted, or overwhelmed by search results that don’t actually help.

Recovery-Focused Design

• Non-clinical but informed: The FAQ is designed to guide people toward better questions and better options, not to replace professional care.
• Context-aware: Understands that “detox”, “rehab”, “aftercare”, and “relapse” have different implications depending on the person and situation.
• Accessible language: Answers are written in plain language while still being precise.

Technical Overview

• Embeddings: Thousands of curated question-answer pairs turned into vectors for semantic retrieval.
• RAG Pipeline: Query → classification → retrieval → rerank → answer synthesis.
• Infrastructure: Vector database + RAG microservice behind a PHP/JS front-end.

Using the FAQ

To try the system live, visit the FAQ interface at:
AddictionTube RAG FAQ

For organizations that want a similar semantic FAQ engine for their own content, see the Vector Search (Unified DB) and Writings & Articles pages for related work.