Welcome to the course: what you will learn

From using AI to building with it. Imagine your manager walks over and asks, can we use AI for this? Maybe it's predicting which customers are about to cancel, or letting staff ask questions about hundreds of internal documents, or sorting a flood of support tickets automatically. Today, a lot of developers and IT people freeze at that moment, because AI feels like a separate world full of math and buzzwords. By the end of this course, you'll be able to answer that question with confidence, build a working solution, and run it in production like any other service you're responsible for. Welcome to Applied AI for Developers and IT Professionals.

Who this course is for. This course is for people who already work with technology. You might be a web or backend developer, a database administrator, a sysadmin, a support engineer, or a data analyst. What you need is some programming or scripting experience and comfort with a terminal. What you don't need is any background in machine learning, statistics or advanced math. Whenever a new term shows up, we'll explain it in plain words, usually with an everyday example like online orders, server logs or support tickets. The goal isn't to turn you into a researcher. It's to make you a practical builder of AI systems.

Your roadmap. The course is organized into six tracks, and each one builds on the one before it. We start with foundations, move into classical machine learning, then deep learning, then large language models and generative AI, and then the engineering needed to run all of it in production. The last track is an optional specialization in GeoAI. Lessons are short, around eight minutes, so you can fit them into a busy week. And most courses end with a project, where you put the ideas together on a realistic problem instead of just watching.

Foundations and classical machine learning. The first two tracks give you the core. In Foundations, you'll set up a proper workstation, get fluent with Python's data tools, learn only the math that machine learning actually uses, and practice preparing messy real world data. Then in Classical Machine Learning, you'll learn the workhorse techniques that still solve most business problems: regression, classification, decision trees, gradient boosting and clustering. You'll also learn how to tell whether a model is actually good. Along the way you'll prepare a customer churn dataset, predict house prices, and segment customers by how they buy.

Deep learning. In the Deep Learning track, we open up neural networks and see how they actually learn, step by step, and then build them in PyTorch. You'll work with images using convolutional networks and transfer learning, which lets you reuse powerful pretrained models instead of starting from zero. You'll also cover time series forecasting and the practical side of training bigger models, like using GPUs, saving checkpoints and debugging a training run that isn't behaving. The track project is classifying product photos, a very common task for online stores.

LLMs and generative AI. The LLM track is where many of you will spend the most time. You'll understand what tokens, embeddings and attention really are, how models like Claude and GPT are trained, and how to call them from code. Then you'll build retrieval augmented generation, usually called RAG, which lets a model answer questions from your own documents. You'll build agents that call tools and APIs, learn the Model Context Protocol, and find out when fine-tuning is worth it. Projects include a chat assistant over internal documentation and an IT helpdesk agent.

MLOps and AI engineering. A model on your laptop doesn't help anyone, so the MLOps track is about shipping. You'll turn a notebook into a clean project, serve a model as an API with FastAPI, package it with Docker, and set up versioning and CI/CD. You'll learn to monitor models once they're live, because data changes and models quietly get worse. The last course focuses on production LLM apps: caching, streaming, observability, automated evaluation, and security issues like prompt injection. If you already run services at work, this track will feel very familiar.

Optional: GeoAI specialization. The final track is an optional specialization in GeoAI, for anyone who works with maps, locations, or satellite and aerial imagery. It starts with a short primer on geographic data, so you don't need any GIS background. Then it applies everything you learned earlier to spatial problems, like detecting changes in imagery or letting people query map data in plain language. If location data isn't part of your work, you can safely skip this track. The core five tracks stand on their own.

How to get the most out of this course. A few tips to get real value from this. First, code along. Pause the video, type the examples yourself, and break them on purpose to see what happens. Second, do the short exercise at the end of each lesson before you move on, because that's where the learning sticks. Third, treat every project as a starting point. Once it works, change the data or the question and make it your own. And finally, a steady pace of a few lessons a week will take you much further than watching ten in one weekend.

Let's get started. So that's the journey. You'll start with solid foundations, learn classical machine learning and deep learning, build real applications with large language models, and finish by running AI in production, with GeoAI as an optional extra. Everything is practical, explained in plain language, and built around projects you can show others. In the next video, we'll take a closer look at the first track, Foundations for AI, and what you'll be able to do by the end of it. Let's get started.