We start with the basics: how AI takes in information and why it answers the way it does. Then we cover how to assign AI a task and check its work, as a manager checks an employee. Depending on participants’ level, we provide our mini-system for working with AI, and the company gains common rules: which data can be shared with AI and which cannot.
A good fit if many of your people have ChatGPT but each uses it differently and the benefit to the company is not visible.
Discuss your task
Over 2–3 sessions, we explain how AI works and teach people to assign it tasks and check the results. Between sessions, participants try applying this in their work. If something does not work, they can bring their own example — we will help identify where the difficulty arose and what to try differently.
Another 2–3 sessions. We provide our mini-system for working with AI and explain how it works. We teach people to develop it independently for their own tasks and share successful solutions with colleagues, so the team can use each other’s experience.
We adapt the number of sessions to the team: there can be more or fewer.
Not the chat buttons, but the principle: how a model takes in information, why it answers the way it does and how to get results from it.
Everything you have given it: the request, files, rules, previous conversation. The main skill is gathering everything it needs so that it performs the task instead of guessing.
A goal, checkpoints, breaking the task into steps; do — check — correct, and repeat. Everything a manager does with an employee needs to be done with AI too.
Give the answer to other AI “critics” for review: one looks for missed rules, another for invented facts. Repeat checks where errors are costly, and make sure a correction does not break earlier work.
Depending on participants’ preparation, we provide our mini-system and teach them to use it. We show how to save a verified solution so AI can repeat the work and colleagues can use your experience.
For advanced users: what parts the system consists of and why it works. Once you understand this, you can break almost any clear recurring process into blocks — from a work report to a family budget.
Employees understand how to work with AI, assign it tasks and check the results. The team knows which company data can be shared with it and which cannot.
Participants in advanced sessions receive our mini-system and understand how to develop it for their own tasks and share verified solutions with colleagues.
Case study diagram · click to enlarge
To get results consistently rather than by luck, and turn a successful one-off chat into repeatable work, so you do not have to explain everything to AI afresh.
AI does exactly what it understood from the task. Like a new employee, it needs a clear goal, interim checks and feedback — otherwise it will fill in the gaps itself.
No. We start with the basics and explain how AI works in simple terms. Participants then try applying what they have learned in their work.
Smart does not mean error-free. We teach people to have results reviewed and repeat the review where errors are costly.
It already has rules that help AI check information and work consistently. You add information about yourself and your work, save successful solutions — and can use them again without explaining everything afresh in every chat.
People who already work confidently with AI and are ready to go deeper. It is not easy, but afterwards they can break their own processes into blocks — at work and at home.
Usually 2–3 for the whole team and another 2–3 for advanced users. We adapt the number to the team.
We show how to explain your style to AI using examples and save the rules for future use. Employees will be able to refine these rules themselves if they are unhappy with the result.
For two weeks, we answer questions in a shared channel and recommend regularly sharing work examples within the company to build a library of verified solutions.
No. After training, people work with AI in work accounts; AI assistants connected to your software and data are a separate next step, if needed.
Hello, I’m Nikita Sumin, founder of Simplit.
I read every message myself and try to reply as quickly as possible. Every new AI task is a genuine interest and a challenge for me. Tell me what takes up the most time in your company, even if you are not yet sure AI is needed here.
A couple of sentences is enough. You can choose a time in the calendar straight away, or message us on WhatsApp or Telegram.
