The development of artificial intelligence is one of Kazakhstan's key priorities: expert opinion from Daulet Ermannov

Daulet Ermanov, specialist in artificial intelligence and machine learning

On the development of Kazakhstan's AI industry

It is currently difficult to imagine a world without artificial intelligence. Kazakhstan's AI industry, in turn, aligns with global trends: there is a focus on the development and application of generative artificial intelligence, the growth of multimodal models, and the expansion of its application capabilities through the use of AI algorithms on edge devices. Notably, this year Kazakhstan developed a concept for the development of artificial intelligence for 2024-2029. This means that the enormous potential of artificial intelligence is recognized in Kazakhstan even at the state level, so in the future this industry could become one of the main areas of development.

As a strength of Kazakhstan in the development of AI technologies, I would note the availability of data, digital potential, and the adaptability of the legal framework to digital business models. Meanwhile, the weaknesses include the absence of large private technology companies, the unavailability of venture capital, and insufficient spending on research and development. Today, not enough people in Kazakhstan have a good understanding of artificial intelligence technologies and their capabilities, which leads to low demand for technological developments and innovations. Kazakhstani companies are also not always ready to invest in innovations and use artificial intelligence.

Currently, the priority sectors of the economy for the implementation of artificial intelligence can be called healthcare, education, logistics, and agriculture. In healthcare, artificial intelligence can be used for analyzing medical data, diagnosing diseases, predicting treatment outcomes, and optimizing the operations of medical institutions. In education, these technologies help automate homework checking, analyze student performance, and personalize the learning process. Logistics also remains one of the priority sectors for the implementation of artificial intelligence. Here it can be used to optimize delivery routes, forecast demand for goods and services, and manage inventory. In agriculture, artificial intelligence is used to monitor crop conditions, forecast yields, and automate agricultural work.

The impact of AI technologies on the startup ecosystem

In Kazakhstan, various events are constantly held that help local startups showcase themselves. For example, UN ESCAP conferences, EUSPN, ICTH, and others. Many startups are involved in the Kazakhstani market, engaged in integrating AI into the aforementioned sectors — these are Cerebra, Ozim Platform, Global Green Hub, Higgsfield AI, Oqylyq, and others. One of the successful cases was the domestic platform Cerebra, which carries out automated stroke diagnosis in accelerated mode. After successful testing in Kazakhstan, the project entered the Uzbek market. Higgsfield AI made a significant breakthrough in the field of AI, creating an advanced system that helps scientists and engineers train complex computer models. Ozim Platform supports people with special needs through adapted educational programs, Global Green Hub works on environmental technologies for sustainable development, Livin creates digital solutions to improve quality of life. Oqylyq helps increase literacy among children, Ziz inc develops smart home solutions with IoT. Among other examples of neural network implementation, one should mention the digital map "Family." It allows providing more than 40 services to socially vulnerable layers of the population, pensioners, and other groups.

Personal experience

The field of technology, in particular artificial intelligence and machine learning, interests me with its transformative potential for solving complex problems and improving life. Working in this field allows me to constantly study and apply advanced methods to solve real-world tasks, which remains both an intellectual stimulus and a professional reward.

In February 2024, we founded TIDAAR AI Solutions company, which focuses on the development of products based on artificial intelligence. The company has two main product lines: the ML Forecasting forecasting tool and the HR Admin face recognition tool. The face recognition algorithm is my personal development. Since I had extensive experience working on computer vision projects and spent a significant amount of time developing custom artificial neural network algorithms, I became an expert in the field of computer vision. My skills in Python combined with CUDA technologies were necessary to create a reliable and efficient face recognition system. All these factors influenced the final decision to create a computer vision product at TIDAAR.

The idea for the HR Admin project arose from my experience in the field of computer vision and the desire to help local companies automate their business processes. A conducted marketing study showed an acute need for a human resources management system that would simplify tracking employee attendance and minimize errors arising from manual data processing. Inefficient attendance management can lead to financial losses, and traditional accounting methods often cause inaccuracies in reports, salary disputes, and uncertainty regarding employee presence at workplaces. My solution is ideal for companies with a large staff, such as shopping malls, manufacturing enterprises, restaurants, and food courts. The HR Admin project is designed to demonstrate how artificial intelligence can facilitate the transition from manual processes to fully automated systems, which in turn increases operational efficiency, data accuracy, and human resources management quality.

The implementation of the project included several key stages. At the first stage, I developed the concept by conducting surveys among business owners to identify areas where computer vision technologies could improve the efficiency of their processes. This data helped me form the idea for HR Admin. The system development took about a year, with significant attention paid to optimizing the face recognition algorithm. The use of CUDA technology for image processing on graphics processors allowed increasing system performance by more than 30 times. After completing this stage, I created a platform that allows users to interact with the HR Admin system. This development stage took about three months. Then I began searching for clients to test the solution. Although there were several offers, I decided to conduct a pilot implementation in one of the shopping malls in Astana. Thanks to HR Admin, the shopping mall now manages personnel much more efficiently, eliminating operational problems. In the accounting department, calculating employee salaries now takes just minutes, whereas previously this process took two to three days.

Since I developed this product independently, I adhered to the standard development process: carefully testing and optimizing it at each stage. During the implementation of the project, a new approach to employee management using face recognition technology was developed and applied. This is the first case of applying such a system in Kazakhstan. It allows for more effective and reliable tracking of employee attendance.

Although there are comprehensive solutions developed by foreign companies, similar systems were not available in Kazakhstan. Foreign solutions are often expensive, and their installation and operation take much longer. My solution is more affordable and can be deployed within a day. In addition, our project is adapted to the specific needs of local businesses. Therefore, the competitiveness of existing companies in the market seemed low to me.

Although the startup was launched only a few months ago, the human resources management system has been working for about two months. We have received requests from various shopping malls across the country, indicating that this project is setting a trend in the market.

The project has had a huge impact on both the development of our startup and on me personally. I received valuable information about the operation of shopping malls in Kazakhstan. These insights inspired several ideas regarding future products and improvements. Moreover, I had the opportunity to establish contacts with major businesspeople in Kazakhstan. This project also provided a platform to demonstrate my experience, which opened up potential opportunities for business expansion.

Further development of the AI industry in Kazakhstan

The main focus will be on developing a modern legal regulatory framework that will be oriented towards world practice and take into account market needs. The first priority will be to increase computing power. Therefore, an increase in the construction of data centers specializing in artificial intelligence is quite likely. To this end, such global players as Amazon, Google, Microsoft, and Citigroup may be attracted. The head of state has repeatedly stated that the venture capital market is planned to be developed as a key driving force of innovation and technological progress, so measures will be taken in the coming years to address the problem of the unavailability of venture capital.

Author: Elizaveta Trudova