Before RAG and MCP, learn the basics

Lots of conversations about RAG and MCP, and I end up falling into a bias. What I see in trainings is that most people don’t realize that knowing the basics is what makes the big difference.

Do you decide which model you’ll use? Do you think about the basic prompting techniques (persona, clear goal, examples, output format, how you want it to think through the problem (CoT))?

After that, it’s really worth using models NOT in ChatGPT or Gemini, but in the OpenAI Platform or AI Studio. Learning to tweak what comes “by default” is an excellent skill: reasoning effort, temperature, top-p/k…

Only then come the tools (and MCP) and RAG.

Understanding how models work, knowing how to choose, and knowing what to put in the input is by far the most important thing.

This is a mental note to myself :)

Originally written in Portuguese and published on LinkedIn on June 30, 2025. I translated it into English with the help of AI (Claude): it let me bring my LinkedIn posts to people who don’t read Portuguese without rewriting each one from scratch. I reviewed the translation to keep my original meaning and tone.