Recently, the Estonian Conference and Training Centre published an interview with me on Delfi about practical AI implementation in business.
During their recent course promotion, someone asked in the comments: "Show us these successful people whose lives AI has changed!"
That's a fair question—and one I hear often. So I've selected the most practical parts of that interview: concrete cases with numbers, timelines, and results you can actually measure. Here are the key questions from the conversation:
What does "implementing AI in business processes" mean? Is it just buying paid ChatGPT for the company?
Implementing AI isn't about buying a subscription—it's about rethinking processes. It's realizing exactly where in our work artificial intelligence can help do something faster, more accurately, and more efficiently.
ChatGPT, Microsoft Copilot, n8n, and dozens of other tools—these are just means. Real value appears when we know how to integrate them into the company's daily life and ensure that employees, clients, or partners actually use them.
You can buy a hundred licenses, but if no one uses them at work—there will be no effect. Therefore, true AI integration is a whole complex of steps:
- analysis and rethinking of business processes,
- selection and configuration of suitable tools,
- employee training and creating internal motivation,
- measuring results and continuous improvement.
It's important to understand: AI implementation is not a one-time action, but a strategic process. Only a systematic approach allows you to turn artificial intelligence from a beautiful idea into a real source of benefit and growth.
Name a project that paid off fastest and that you remember most. What were the initial KPIs, what did you get, in how many weeks?
Probably the most memorable case happened right during one of our AI implementation workshops at a large factory in Ida-Virumaa.
A company with multi-million turnover... and their website was created 25 years ago—basically a copy of a Wikipedia page.
During the practical part, I simply photographed their old paper brochures, uploaded the texts to ChatGPT, asked it to rewrite them in an understandable and modern form, and then through another tool generated a new, visually neat website for them in 18 minutes.
For the participants, this was a real wow moment. They literally saw before their eyes how AI can transform in minutes what previously seemed complex and expensive.
Of course, most projects aren't done that quickly—full-fledged solutions usually take 2 to 6 months, depending on their scale and complexity. But this case became an excellent illustration of how powerful the effect of proper AI application can be—even in a seemingly traditional industry.
Do you have specific numbers: how much the company invested—how much they got out? Name a real case where AI gave profit/savings. What were the "before/after" numbers?
Yes, of course. We've already accumulated quite a few such examples. One of the most indicative is related to a construction company specializing in concrete work.
The client implemented an AI-based solution to automate the procurement process: supplier search, preparation and distribution of price requests, collection and analysis of commercial proposals.
Development and implementation of the system cost approximately €25,000. The result—labor costs for the entire process were reduced by almost half.
If before the project they needed three people on a permanent basis, after implementation one employee manages, with support from another in an auxiliary role. Other specialists were transferred to the project management department.
The project lasted about four months and paid off in six months. From the company's perspective, this is not only savings, but also growth in productivity and process transparency—an effect that continues to this day.
Did you have to advise a client about specific position cuts in the company? What tasks went to machines, what skills became critical for humans?
With one client, we conducted a project to implement AI in the sales process. Out of ten employees, two—AI ambassadors—began actively using ChatGPT and several auxiliary tools at every stage of the deal: from preparing proposals to communicating with clients.
After two months, their productivity practically equaled the productivity of all the other eight colleagues combined.
After that, the company revised the entire sales process—now using AI became a mandatory part of work for all managers. No one was fired, but roles and requirements changed: now not so much mechanical actions are important, but the ability to work with data, interpret information, and formulate queries for AI.
In fact, the salary range remained the same, but it's now justified not by experience, but by the ability to effectively use technology.
If you were forbidden to use the words "efficiency" and "savings," how would you measure AI's value? Give 3 metrics, not about money.
If we remove "savings" and "efficiency," then for me AI is primarily about growth.
Growth of intellectual capabilities of people and organizations. Access to practically unlimited knowledge in seconds. The ability to apply this resource to grow as a company, as a professional, and as an expert.
Therefore, if measuring AI's value with non-material metrics, I would choose these three:
- Competency growth. How much people began to understand more, deeper, and act more consciously.
- Growth in learning and adaptation speed. How much faster a company or specialist can master new topics and technologies.
- Growth in decision-making confidence. When you have a tool that helps you think, test hypotheses, and make decisions—the level of confidence and quality of these decisions noticeably grows.
AI, in my opinion, should not just reduce costs, but help people and organizations become smarter and stronger.
Why do companies resist AI implementation? What are they afraid of?
Fear is an absolutely natural reaction, especially in companies that previously invested little in technology.
We often work with organizations where part of the processes are still conducted in Excel or even on paper. Such companies objectively haven't yet completed the full path of digital transformation, and for them, AI implementation looks like too big a risk.
The main reason for resistance is fear of wasting money. Managers understand that technologies are rapidly developing, but aren't always confident they'll be able to get tangible returns on investment.
And this is essentially justified fear. Experience working with AI cannot simply be "bought"—it needs to be earned. Like in sports: you can watch hundreds of videos about technique, but until you go out on the field, there will be no progress.
Therefore, I always tell clients: a small but real step in AI gives more than a perfect plan postponed for later.
If a company doesn't have an AI strategy, where should they start?
This is quite a typical situation—companies want to use artificial intelligence but don't know where to start. Therefore, I usually propose moving through three sequential steps.
Step 1. Define strategic goals and challenges
AI isn't implemented for the sake of implementation itself. Its task is to help the company achieve specific goals: sell more, work faster, improve customer experience, make more accurate decisions. Therefore, the first step is to clearly understand where exactly technology can become a tool for achieving key business results.
Step 2. Study the technology and its real capabilities
AI isn't magic, but a specific set of technologies with clear boundaries of applicability. At this stage, it's important to form a common understanding within the company of what exactly AI does, where it's useful, and where it's not.
For this, we conduct workshops and trainings where participants see in practice how modern tools work and what tasks they can solve today.
Step 3. Start with 2-3 initiatives where you can quickly see results
Initially, it's important not to build large-scale systems, but to choose specific, understandable use cases where AI can show real benefits without large investments. These can be processes related to information and document processing, communication automation, or implementing tools like Microsoft Copilot and ChatGPT into daily work.
Such "pilot" steps help calculate ROI, see value, and build trust in technology within the company.
Approximate budget
For most organizations, the first stage of AI implementation costs between €5,000 and €30,000, depending on company size and depth of involvement. Usually, this is enough to conduct training, workshops, and implement first pilot projects. First results become noticeable within 1-3 months.
How to understand that implementation is actually working
There are three simple signs:
- Employees begin using AI tools in daily work
- The company sees measurable effect—time savings, improved quality, or task execution speed
- Interest and confidence to continue the path appears internally—AI stops being an experiment and becomes part of corporate culture
This is an excerpt from a longer interview. Read the full version with additional case studies and implementation details on Delfi (in Russian).
Ready to Explore AI for Your Business?
If you're wondering how AI could work specifically for your company—let's talk. We offer consultations where we analyze your processes, identify opportunities, and create a practical roadmap for AI implementation.
Book a free consultation to discuss your specific case and see what's possible.