GeoStorm, a startup with Ukrainian roots, develops tools for accurate weather forecasts

The company GeoStorm Technology, with offices in Israel and Ukraine, is developing accurate weather forecasting technology. Recently, the startup’s team joined the MassChallenge accelerator in Boston, and now it is working on raising investment.

AIN.UA’s editor tells the project’s story.

What does GeoStorm do?

GeoStorm Technology, founded in 2020, spun off Carpe Diem Solutions, an Israeli developer of IoT solutions, with Ukrainian and Israeli co-founders. The company was started by Oleg Dryzhak, Gennadiy Zhuga, and Andriy Semeniaka (the latter two co-founders are senior researchers and Radio electronics specialists at Kharkiv National University of Radio Electronics, NURE). It has offices in Tel Aviv and an R&D center in Kharkiv. The team comprises 13 employees.

The project’s co-founders

What do they work on?

The company operates in the field of climate forecasting and control. The team is developing a technology that enables it to recognize real-time extreme weather activity.

According to the developers, the accuracy is quite high: the speed of a weather formation’s movement is read to within 0.1 m/s, and the critical parameters surveillance time is within 1 minute.

The company’s representatives explain that during 2019 only, climate change caused 409 natural disasters globally, costing $232 billion of damage. Many large high-tech companies keep an eye on the advances in this field.

Google has revealed its plans to launch technology with new AI models, which allows forecasting the weather instantly. IBM has announced a service called IBM Weather Signals, based on its own AI system Watson and designed to aid business planning.

According to Oleg Dryzhak’s comment to AIN.UA’s editors, the company is developing and selling a digital signal processing (DSP) module with ALFA algorithms installed. Such modules are usually installed on radars designed to measure the weather at weather stations, maritime vessels, air services, etc.

Such algorithms provide for:

  • the filtering of noises which prevent accurate precipitation and wind speed readings (for example, interference from 5G and LTE towers);
  • faster radar data processing;
  • exact weather event mapping: for example, such data would allow a farmer to determine which of his or her fields could be threatened by hail, and which not.

“The weather forecasting is carried out using complex computing models and supercomputers receiving data from radars and processing them for up to 6 hours. And we do it just immediately on the motherboard, modernizing the customer’s radar,” one of the co-founders explains.

Data from these radars are processed by artificial intelligence systems, also developed by the startup’s team.

How do they plan to generate profits?

According to the project’s co-founder, the first income of the company is selling DSP modules with algorithms that are integrated into a customer’s software and supercomputers (if any).

The team applied for participation in the accelerator MassChallenge, taking into account the results of its graduates: 89% are still active or have made an exit, more than $6 billion of raised investments, $3 billion in revenue.

“The goals we set for the period of the program are to check the value offer and the business model, develop sales and partnerships, attract experts to the team,” says the co-founder.

Currently, the startup conducts pilot tests of its technology with weather stations in Israel and Europe, and the ClimaCell weather platform, as well as negotiates with radar manufacturers.