RAG & AI Agents Course for Developers: Vector DBs and Automation
Build a real retrieval-augmented chatbot from scratch: embeddings, pgvector search, a free-tier LLM, n8n automation, and custom agents -- using this site's own live chatbot as the worked example.
About this course
A hands-on course covering generative AI fundamentals, retrieval-augmented generation (RAG) architecture, vector databases (with a deep dive into pgvector), setting up a capable LLM without paying for API access, automating AI pipelines visually with n8n, and building custom tool-calling agents. The capstone walks through exactly how this portfolio's own AI chatbot widget is built end to end -- the same embeddings, the same vector search, the same free-tier model -- so every concept is tied to a real, running system rather than a toy example.
Try the first lesson free
Read the opening of the course in full, no account needed.
Course index
- What "Generative AI" actually means for builders
- RAG architecture: the four moving parts
- Vector databases and embeddings: what "similarity" actually means
- Toy retrieval lab: build intuition for "search" without a real model
- Setting up a capable LLM without paying for one
- n8n: automating the pipeline around your LLM
- Custom agents: giving the model tools, not just text
- Capstone — how this site's own chatbot is actually built
- Where to go from here
Ends with a final quiz. Score 70% or better and you can download a completion certificate in your own name — shareable on LinkedIn and verifiable by its certificate code.