Kazakhstan modernises its statistical system with AI and new data sources

In Kazakhstan, Statistics Workers’ Day is celebrated annually on November 8. Over the course of its history, the sector has evolved from paper-based reporting and manual calculations to digital platforms, artificial intelligence and big data analytics.

This year marks the 106th anniversary of Kazakhstan’s statistical system. Today, official statistics serve as one of the key instruments for monitoring and analysing economic and social developments in the country.

As part of the Year of Digitalisation and Artificial Intelligence, the Bureau of National Statistics continues to introduce technologies that accelerate data collection and processing and improve access to statistical information. Key achievements include the modernisation of the E-Statistics system and the Respondent’s Personal Account, the development of the National Reference Information Service, and the launch of an upgraded version of the Taldau platform featuring the AI Taldau intelligent assistant.

Today, 97% of primary statistical data are submitted electronically. The use of big data is also expanding.

The reporting burden on enterprises and organisations has been reduced by 30%. Much of the required information is now obtained from existing information systems, eliminating the need for businesses to complete redundant forms.

Less Manual Work

One of the major initiatives has been the modernisation of the E-Statistics system. It integrates all stages of statistical production, from planning and data collection to processing. Processes have been automated and aligned with a unified schedule, reducing manual operations and errors while accelerating the production of statistical results.

Submitting statistical reports has also become easier. The upgraded Respondent’s Personal Account offers a more user-friendly interface, enabling enterprises to complete and submit statistical forms more quickly, with many operations requiring just a few clicks.

To facilitate data exchange between government agencies, the Bureau has developed the National Reference Information Service. It enables all government information systems to use common reference databases and classification systems. This allows public authorities to exchange comparable data and avoid collecting information that is already available. Official statistics are currently compiled using data from 36 administrative sources and four alternative data sources.

Receipts and Mobile Network Data

New technologies are also transforming the ways in which statistical information is collected. Data generated through everyday purchases and travel complement traditional statistical surveys and provide a more comprehensive picture of ongoing changes.

One example is the Consumer Price Index (CPI), which is now also calculated using fiscal receipts. The total volume of such data amounts to 8.3 billion records.

For CPI calculations, approximately two million price observations are selected from nearly 15,000 retail outlets. Since the beginning of this year, prices for 120 goods have been calculated using receipt data. Artificial intelligence helps automatically classify goods into relevant categories and process large volumes of information more efficiently.

Tourism statistics also make use of anonymised mobile phone location data. These data are used to assess tourist flows, popular destinations, length of stay and seasonal fluctuations. This approach provides a more accurate understanding of tourism development across the country’s regions.

Statistics at the Click of a Button

Alongside improvements in data collection methods, the Bureau is expanding its services for users. The upgraded Taldau platform provides a single interface for searching statistical indicators, comparing data across regions and time periods, generating charts and maps, creating analytical dashboards and downloading information in various formats.

The platform’s key new feature is AI Taldau, an artificial intelligence-powered assistant. Users can simply enter a question in natural language. The assistant retrieves the relevant data, presents them in tables and charts, compares indicators and prepares analytical materials. Users can also verify the underlying sources and calculation methodology.

Artificial intelligence is being integrated into other Bureau services as well. The AI Stat assistant facilitates information searches on the official website, while a chatbot powered by the statistical business register provides, through a conversational interface, information on legal entities registered in Kazakhstan, their branches and individual entrepreneurs. During the 2025 Agricultural Census, participants’ questions were answered by the Sanaq Myrza chatbot.

“Amid digital transformation, the very philosophy of statistical production is changing. The introduction of artificial intelligence, process automation and the use of new data sources enable the Bureau to improve its engagement with respondents, minimise their reporting burden and provide users with reliable official statistics in a timely manner,” said Maksat Turlybayev, Head of the Bureau of National Statistics.

The digitalisation of statistics forms part of Kazakhstan’s broader efforts to develop a data-driven economy. An important milestone in this process was the adoption of the Digital Code, which establishes the regulatory framework for data circulation, IT product development and the operation of information exchange platforms. Since 2024, the Bureau has been cooperating with the National Bureau of Statistics of China and the Shanghai Data Exchange.

The next stage will involve the wider application of artificial intelligence in statistical production itself. Priority areas include the automated detection of errors and anomalies, data validation and classification, and the preparation of analytical outputs. This will enable specialists to spend less time on routine tasks and focus more on analysing results.