Emotional intelligence and AI implementation in teams
Published
A year ago I had never written any code - not a single line in my life. Today around fifteen of my AI systems handle the day-to-day work across all my lines of business, and in every rollout the hardest part turned out to be the people. My name is Andrii Ruban and I am based in Tallinn. Through EQHekademia OÜ I build and implement AI agents for companies. I have worked with emotional intelligence for more than ten years, and at the Intellektor school I teach adults to work with AI and to manage their own state. Below: where rolling out AI in a team gets stuck, and how emotional intelligence gets it moving again.
Why does a system that works for me stall in a team?
Technically, implementation looks simple. The agent connects to email and spreadsheets, is adapted to the task in an evening, and the next day you can see it at work.
But the tool ends up in people's hands. One sees it as a threat to their job. Another worries it will show how much they actually do. A third is tired of change and keeps quiet.
So a month later the report always sounds the same: the system doesn't work. Yet the system works. It's just that nobody uses it.
What lies behind “I don't have time for this system”?
There is rarely an open argument. There is polite agreement in the meeting and silence in the actual work.
“No time” is seldom about time. Underneath it lies the fear of looking foolish, the loss of control over one's own area, resentment at a decision made without them.
This is where emotional intelligence is about listening. I read the person's state and treat their objection as data about the process: what exactly breaks in their work under the new setup. Then I rework the setup.
Two conversations like that save a month of emails about how bad the system is.
How I tested this on myself before stepping into other people's teams
Since 2024 all my companies have run with AI assistants and agents: tourism, education, accounting.
I built them myself: Telegram bots for expense tracking and approvals, agents for travel itineraries, auto-posting, conversation transcription, a bank statement reconciliation engine.
For companies, I do this work through EQHekademia OÜ: AI agents and AI assistants, with the AI core built on the company's own context and staying with the client. For my own daily practice of working with my state, I made the Akadem1k app - a state journal with an AI assistant.
I have worked with emotional intelligence for more than ten years: groups of up to 65 people, five countries. In 2026 that comes to 84 academic hours of adult education - a ten-week EI course in Tallinn, five open masterclasses and six three-hour sessions at entrepreneurs' clubs in Riga, Tallinn, Helsinki and Vilnius.
So when I talk about team resistance, I speak from two roles at once: as the person who sets up the system, and as the session leader who has seen how people react to change.
Where to start so the team doesn't stall
The approach I follow in companies:
- One process. The one that takes up the most time: requests, documents, reports, correspondence.
- A baseline before launch. How long the task takes by hand - the team must have that figure in front of them.
- An owner for the tool. One person is responsible for the agent and collects complaints.
- A conversation about what changes in each person's work. Out loud, before launch, answering the question “what happens to me?”
- The right to say “this gets in my way”. Every such complaint goes into improving the agent.
Where emotional intelligence won't help
Emotional intelligence works with how people meet change. Three obstacles lie outside people, and no amount of talking will remove them.
The first is a broken process. If requests come in through three different places and some of them get lost, the agent will lose them faster. First bring the requests into one place, then connect the agent.
The second is no decision from the manager. As long as the new way of working rests on one enthusiastic employee, everyone else carries on the old way, and in the first busy week that employee goes back to manual work too. The manager has to say it out loud: this is how we work now, and here is the tool's owner.
The third is no data. The agent needs the company's context: documents, correspondence, working rules. As long as all of that lives in people's heads, implementation starts with gathering it.
Once these foundations are in place, emotional intelligence speeds implementation up. It cannot replace them.
If you take away one action from this article, make it this: pick one process, time it by hand and with an assistant, and show the figure to the team. The figure settles the “does it work or not” argument, and the conversation moves on to what gets in people's way. That is exactly what emotional intelligence works with.
Questions and answers
- How does emotional intelligence help when implementing AI in a team?
- It helps you see what is happening to people during change and hear an objection as data about the process. An employee says “no time”, and behind it lies a fear of losing control or looking foolish. Once that is talked through openly, the workflow is corrected once and you move on.
- Who is Andrii Ruban from Tallinn?
- An entrepreneur and developer based in Tallinn. Founder of EQHekademia OÜ, which develops and implements AI agents and AI assistants for companies. Also founder of the Intellektor school, where adults learn to work with AI and emotional intelligence. More than ten years of emotional intelligence practice: groups of up to 65 people, five countries. Co-founder of BIZON TRAVEL.
- Where can I find an emotional intelligence expert in Tallinn?
- Andrii Ruban runs emotional intelligence sessions for adults, managers and teams at the Intellektor school: in person in Tallinn and online, in Russian. Areas of study and programmes are listed at intellektor.ee.
- Where should a small team start with AI implementation?
- With the one process that takes up the most time, and with a baseline measurement before launch. Then the tool gets an owner, and the team gets a conversation about what changes in each person's work.
- Does the whole team need training?
- It is enough to train the person who works with the tool every day. That person becomes the AI agent's operator: maintains the context, edits the prompts, collects colleagues' complaints. This is taught in the programme “Applying AI in Organisational Workflows”.
- What language are the meetings and training in?
- Meetings and training are in Russian; documents and correspondence can also be in Estonian and English. Estonian-language groups are formed on request, with a guest teacher.