This course contains the use of artificial intelligence.

Most AI pilots succeed. That is the problem. A pilot proves the technology works with three enthusiastic volunteers, and proves nothing about whether four hundred people will change how they work.

Why adoption stalls after the pilot

The tool was rolled out to a process nobody had described, so it now accelerates whatever was already confused. Training was a one-hour session and a recording, after which everyone returned to the way they knew. The people who benefit most are the ones with least time to learn. And the resistance is read as technophobia when it is usually about status — a person who was valuable because they knew how to do something quickly, in a world where it is now done instantly.

Then someone asks for the return on the investment, and nobody set a baseline before starting.

How the course is built

Forty lessons in five parts. Where AI actually works first, twelve lessons across a whole function: selection, search, onboarding, evaluation, brand, training, off-the-shelf platforms, rewards, culture, predictive analytics without code, and document and process generation. This is the map of what is realistic today rather than what is demonstrated in keynotes.

Then four lessons on getting usable output: the seven blocks of a prompt, chain-of-thought and tree-of-thoughts, the shift to context engineering with Projects and RAG, and agentic workflows with MCP — including where a human still has to approve.

The process, the people and the change

Then eight lessons on processes, which is the step organisations skip and pay for. Waste, standardisation, maturity, value stream mapping, and the toolkit from 5 Whys to SIPOC. Automating a described process compounds; automating an undescribed one accelerates confusion.

Then ten lessons on teaching people, taken seriously. Adult learning, ADDIE, diagnosing where the gap actually sits, designing the programme, the transfer problem — sixteen factors that decide whether anything is used the following week — post-session support, and honest evaluation.

And finally change management: Kotter's eight steps, the rider-and-elephant model for the emotional half, quick wins, and fixing the change into culture so it survives the person who championed it.

A note on the examples

The twelve adoption lessons run on a single function end to end — people operations — rather than sampling one example from each industry. That is deliberate. Watching AI applied to one function completely, from first pilot to full coverage, teaches the sequencing better than a tour of scattered use cases, and the sequencing is what this course is about.

Who is teaching this

Mike, the number one HR instructor on Udemy. More than 1.6 million course enrolments, over 150,000 professionals trained, PHRi and SHRM-CP certified, HRCI representative in more than 10 countries. I built the people function of the unicorn Preply and worked at Wargaming, Alfa-Bank and iDeals.

Before the rollout

Pick the process you most want to automate and write down how it works today, step by step, including who does what. If that document is hard to produce, automation will not help yet, and knowing this before spending the budget is most of the value here. Enrol now and start the first lesson today.