From Zero to RAG Agent: Full Beginner's Course (no code)

From Zero to RAG Agent: Full Beginner's Course (no code)

by Nate Herk

Applied Sciences & Engineering AI — Models & Research UYQArtificial intelligence (AI)UYQMMachine learning
🎙 Nate Herk 👥 964K 📅 July 21, 2025 ⏱ 22 min 👁 163K 📄 tutorial 🧭 2026-08-28
Available in: English (current) Français

Keywords

RAGvector databaseembeddingsn8nSupabase

Summary

This tutorial guides complete beginners through building a Retrieval-Augmented Generation (RAG) agent without writing code, using n8n and Supabase. The video begins by explaining RAG as retrieving information to generate accurate answers, clarifying that it is not limited to vector databases. It then introduces vector databases, illustrating how text chunks are embedded and placed by meaning. The RAG pipeline is broken down into chunking, embedding, and vectorizing a document, using a 22-page PDF on golf rules as the example. The creator demonstrates setting up a Supabase project, creating a vector store table via SQL, and connecting n8n nodes for Google Drive, document loading, OpenAI embeddings, and Supabase vector store. After populating the database with 77 vectors, the video shows building an AI agent with a chat trigger, OpenAI chat model, and a Supabase vector store tool. Testing the agent with a question about practice rules yields accurate answers, verified against the PDF. The tutorial also covers reading agent logs to understand the retrieval process and setting up memory using PostgreSQL in Supabase to store chat history. The video concludes by mentioning extended workflows and additional resources, emphasizing the simplicity of the no-code approach.

193 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video provides high practical value for beginners, offering a complete, actionable walkthrough. The argumentation is clear and logical, building from basic concepts to a working system. The creator effectively uses analogies (e.g., Googling an answer) and visual aids to explain abstract ideas. However, the content lacks critical discussion of limitations, such as retrieval quality, embedding model choices, or potential pitfalls, which would strengthen the argumentation for real-world applications.

Scientific Rigor, Source Quality, Title Accuracy

The tutorial is methodical and accurate in its instructions, with steps that are reproducible. Sources are primarily tool documentation (Supabase, n8n, OpenAI) and the creator’s own community resources, which are appropriate for a tutorial but not exhaustive. The title accurately reflects the content, and the video stays on topic throughout. No external scientific sources are cited, which is acceptable for a practical tutorial but limits the depth of theoretical grounding.

154 words

Title / Content Match

The title accurately reflects the content: a beginner-friendly, no-code tutorial for building a RAG agent.

Quality & Reliability

7/10

Clear, step-by-step tutorial with practical demonstrations. Concepts are accurately explained at a high level, but lacks depth on limitations and alternative approaches. Sources are limited to tool documentation and community resources.

Chapters

Cited Sources

Concurring Sources

Contribution & Novelties

The video offers a clear, no-code entry point to RAG, demystifying the technology for beginners. Its main contribution is the practical, step-by-step integration of n8n with Supabase, showing how to build a functional RAG agent without programming. It also highlights the use of PostgreSQL for memory, adding a layer of persistence. However, it does not introduce novel concepts, as RAG and vector databases are well-established. The value lies in accessibility and clarity.

Pour aller plus loin :

135 words

Radar Profile

The radar profile shows balanced scores across information quantity, quality, and reliability, with a lower technical depth. This reflects a tutorial that is comprehensive for beginners but does not delve into advanced technical details, making it accessible yet not exhaustive.

Reliability 7/10

💬 Très positif. Sur les 30 commentaires analysés, les utilisateurs expriment une gratitude marquée et une forte appréciation pour la clarté et l'utilité du tutoriel, avec de nombreuses demandes pour des contenus plus avancés.