AI
AI
A course for complete beginners. You do not need to know programming. If you can write an email, you can do this.
Course guide · 2026
This guide teaches four things: coding, accounting, security and design — all using AI, all without a technical background.
Everyone begins with the safety rule and the foundation. After that you pick the track that matches your job.
By the end you will be able to ask an AI for help and get something useful, build a small working tool for your own job, analyse your spreadsheets by asking questions in ordinary language, and — most importantly — recognise when the AI is wrong.
An AI is not inside your organisation. Treat it like a stranger on the telephone.
Before you type anything, ask yourself: would I read this out loud to someone outside?
Never share:
Safe to share: general questions, published documents, made-up example data, and your own text with the names removed.
Learn one habit: replace real details with fake ones before you ask. "Customer Hakobyan owes 4.2 million" becomes "a customer owes 4.2 million". You get the same answer. The secret stays where it belongs.
A weak request gets a weak answer. Four things make a request strong:
Try it: write a vague request, look at the answer, then rewrite it with those four points and compare.
Three sessions, for everybody.
1 · What AI is, and what it is not. An AI is a very good assistant and a very bad authority. It writes confidently even when it is wrong, and it can invent a law, a number or a source that does not exist. Your exercise is to ask three questions in your own field and find an answer that is wrong. That is the point of the session.
2 · How to ask well. Turn a vague request into a clear one using the four points above.
3 · Safety rules and the workspace. Log in, find your way around, and practise the data rule until it is automatic.
For anyone who repeats the same manual task, or wishes a small tool existed. You will not learn a programming language — you will learn to describe what you need, let the AI build it, and then check it.
For accounting, reporting, audit and analysis — anyone who lives in spreadsheets. This is where AI saves the most hours immediately.
How your data is handled. Your spreadsheets are read on your own computer. What is sent to the AI is only the file names, the sheet names, the column headings and the first five rows. Everything after the fifth row never leaves your machine.
Be aware of that limit: five real rows are five real records. For sensitive registers, work on a copy where the first rows have been replaced with dummy values.
Always check the arithmetic. AI can misread a column or silently skip empty cells. For anything that leaves your desk, verify the totals by hand the first time.
For everyone, and essential for IT, risk and compliance staff. Two halves: protecting your organisation from careless AI use, and using AI to strengthen security.
For communications, publications, HR, and anyone who makes slides or documents.
Six weeks, two sessions a week, 90 minutes each, in groups of 10–12. You need a computer and a browser — there is nothing to install, and the platform is available in Armenian, Russian and English.
Every session has the same shape: we show you one real example (15 minutes), we do it together step by step (45 minutes), then you do it alone with your own work task (30 minutes).
To finish, you submit one real piece of work built during the course and answer five questions on the data rules. The data questions must all be correct — that is the part that protects your organisation.
AI will not replace your judgement, and it is not trying to. It removes the slow, repetitive part of your work so your attention goes to the part that needs a person.
The people who benefit most are not the technical ones. They are the ones who know their own job well enough to spot immediately when an answer is wrong.