AI has quickly become part of almost every developer conversation. We hear about language models, tokens, prompts, embeddings, vector databases, RAG, agents, MCP, and a long list of other terms. But before we can build useful AI-powered applications, we need to understand what these concepts actually mean.

In this session, we will start from the beginning. No prior AI knowledge is required.

We will look at what a language model is, how prompts work, what tokens are, why embeddings are useful, and what RAG can help us with. We will also talk about "newer" concepts like MCP and agents, without assuming that everyone already knows what they are.

This session is not a deep dive into machine learning theory. Instead, it is a practical, developer-friendly introduction to the building blocks of modern AI applications. Along the way, we may build a few small examples to make the concepts easier to understand.

The goal is to give you a solid foundation, so the next time someone throws words like embeddings, RAG, agents, or MCP into a meeting, you do not have to nod politely and Google it afterwards.