It won't replace power engineers, but it will save money. How DTEK and MHP Eko Energy use AI

DTEK has been interested in artificial intelligence since at least 2018, and it has saved MHP millions of hryvnias, the co-founder of the startup with which the company cooperated told us.How AI is changing the Ukrainian energy sector, what tasks businesses are implementing this technology for, and how much it costs – found out LIGA.net.What MHP Eko Energy says"First of all, the demand for artificial intelligence comes from projects aimed at working with the end user," confirms Marina Bautina, Senior Artificial Intelligence Consultant at SoftServe.In Ukraine, the simplest and cheapest products such as Gemini, ChatGPT, and Copilot are often used. Users in the energy sector include MHP Eko Energy, which supplies electricity and natural gas to consumers in several regions. In 2025, the private joint-stock company earned UAH 3.6 billionand profit amounted to UAH 31.5 million.For example, the company has a "virtual power engineer assistant". It allows us to forecast electricity consumption and optimize the cost of electricity production at our production sites."First of all, we are talking about its use for trading – purchasing electricity and other energy resources on trading platforms and exchanges," explained Oleksandr Dombrovsky, President of MHP Eko Energy."The energy sector is all about numbers: consumption, procurement schedules, optimization, cost of energy resources, etc. Thanks to the use of artificial intelligence elements, we are able to process large databases. In addition, our specialists can choose the most efficient algorithms and choose the best possible solution compared to those that were available a few months ago," adds the head of the company.Read alsoHow to make money on energy storage in Ukraine: experience of DTEK, OKKO and Elementum EnergyHowever, Dombrowski considers AI a new tool in the Ukrainian market, so the company is still studying its effective use. Sometimes AI makes mistakes, so we have to work with several products like Gemini and ChatGPT. They run the same task and then compare the results.Currently, the company sees no reason to spend significant funds on such software products. They are content with the cheapest options.How much does in-house development cost"This is a perfectly normal solution," says Marina Bautina from SoftServe, "For example, a subscription to ChatGPT costs $20-30 per month. And thanks to this, businesses get access to the functionality right out of the box. It makes no sense for a company to pay for the development of a similar solution. Unless the business has certain special security practices."At the same time, some companies invest heavily in both the development and support of the solution, says Bautina. The amount of money spent on development depends on many factors, but can range from 30 to hundreds of thousands of dollars.Read alsoA new era of efficiency: how Ukrainian business uses AIHowever, less money is needed to support the solution. According to Maryna Bautina, it depends on whether the company has its own infrastructure or simply relies on third-party providers. "Therefore, the amount can range from several thousand to tens of thousands of dollars," the expert adds.How AI predicts consumption for DTEKOne of the Ukrainian energy companies that actively invests in AI technologies is DTEK Holding. The company is actively integrating artificial intelligence into its business with the help of its own IT company Modus X, which has already launched a special Quantum program for this purpose.Thus, DTEK plans to optimize logistics, finance, HR, cybersecurity, receive forecasts of electricity production, warnings of possible breakdowns at thermal power plants, etc.In 2023, DTEK acquired FSight from the Israeli company Predict+ AI technology. It helps minimize the effects of missile strikes on the power grid. For example, artificial intelligence is involved in managing and redistributing capacities to reduce spikes in electricity consumption due to arrivals.In 2018, together with Radar Tech, DTEK launched an accelerator to find innovative developments in the energy sector. And in 2019, D.Trading, which belongs to DTEK Group, signed a cooperation agreement with A-Gnostics. Their product predicts energy consumption using artificial intelligence."The price of electricity is different in every hour. Therefore, you can save a lot of money if you consume less during peak hours when electricity is more expensive," says LIGA.net andriy Starzhynsky, co-founder of A-Gnostics. He also added that the use of forecasts saved MHP UAH 2 million every month.Read alsoIt senses the approach of accidents and saves electricity: AI from Ukrainian a-Gnostics enters new marketsExperience of Scandinavian countriesThe example of Scandinavian countries shows how energy companies and consumers can cooperate with artificial intelligence.Sweden's Vattenfall uses advanced machine learning algorithms to analyze data from sensors on wind turbines. AI collects and processes information on vibrations, temperature, rotation speed, and pressure. It then identifies anomalies that may indicate an impending failure of a particular piece of equipment and notifies the company weeks or months in advance.As a result, Vattenfall has reduced downtime for routine, often unnecessary inspections and also reduced the cost of emergency repairs. Thanks to AI, the company can schedule repairs for periods of, say, low wind. Thus, such outages have minimal impact on electricity production.The Norwegian renewable energy company Statkraft has created a "virtual power plant" (Virtual Power Plant). This is a software platform that uses AI models and predictive analytics to combine and balance hundreds of different energy sources: hydroelectric power plants, wind and solar farms, and large industrial consumers. AI predicts the generation of each asset and market demand, and then automatically manages these resources to trade electricity on European spot markets.As a result, Statkraft is able to respond quickly to price fluctuations in the markets by selling surplus energy or buying energy at the best prices. The company has also achieved grid flexibility by making good use of variable energy sources such as solar and wind, as well as efficient use of hydro resources. AI helps to optimize the use of hydroelectric reservoirs, deciding when it is better to generate electricity and when to save water for production during peak demand periods.Artificial intelligence also helps household consumers in Norway save energy and thus money. The latter receive a daily projected cost of electricity for the next month from their electricity supplier.Essentially, it is the use of AI in the operation of the so-called Smart Grid. In Ukraine, the use of Smart Grid is still at an early stage, but work in this area continues despite the war.At the same time, Bautina notes that many Ukrainian companies are still interested in AI mainly because of the hype around it. Instead, she advises starting from a specific problem to be solved. Perhaps simpler tools than AI will be enough for this purpose.

It won't replace power engineers, but it will save money. How DTEK and MHP Eko Energy use AI

Comments (0)

What we fight with: weapons that strengthened Ukraine's defense in 2025

IT professionals in Israel increase requests for relocation abroad due to war – report

Contract from a Turkish prison: How Poltava merchants circumvented Russian obstacles to trade

"Waiting for liberators." The acute problem of "zhduns" in the tactical rear of the Armed Forces of Ukraine

A future without mines: How Ukraine is building a unique humanitarian demining system