Ukraine transformed its State Statistics Service from a legacy institution into a modern, AI-powered data organisation, making official statistics more accessible to government, businesses, researchers and the public.

The Challenge: Outdated Systems Meet Modern Demands

Ukraine’s State Statistics Service (Ukrstat) operated on fragmented, legacy infrastructure. Data was often collected manually, processed across 45 separate systems, some of them 20–25 years old, and published through a static portal. One critical system was maintained by a single specialist, creating significant operational risk. The core stages of statistical production – collection, processing and dissemination – were disconnected, with manual transfers between Excel tables and other systems. Even a simple administrative change could require updates across dozens of separate databases.

The consequences were systemic. Policymakers lacked a modern, reliable data foundation, while international organisations could not easily access Ukrainian statistics in formats compatible with global standards – a critical gap as Ukraine pursued EU integration.

Transforming the institution required more than new technology. With approximately 6,000 staff before 2022, Ukrstat’s workflows were shaped by analogue processes and a traditional, compliance-oriented operating model. Moving towards a more product-oriented, user-first approach required changes to organisational culture, roles and capabilities.

The full-scale war that began in February 2022 added further pressure. Traditional data collection was disrupted, while demand for timely and high-quality statistics increased as Ukraine responded to the economic, social and reconstruction challenges created by the war.

The transformation was therefore both a technological and institutional challenge: Ukrstat needed to modernise how it produced statistics while building a more resilient organisation capable of operating under prolonged disruption.

The Solution: Sigma As A Unified Statistical Ecosystem

The solution was Sigma, a unified technological ecosystem designed to replace 45 legacy systems and support the complete statistical lifecycle. The core of the new system is built on EPAM’s adaptable QuantHub platform, configured specifically for Ukrstat’s needs. Its publicly accessible portal launched in beta in August 2023 and moved into full production in November 2025.

Built on SDMX standards, Sigma supports interoperability with the global statistical ecosystem, including Eurostat, the IMF, the World Bank and the OECD, and provides access to more than 2,000 statistical indicators.

The platform combines several capabilities. StatGPT provides a natural-language interface to official statistics through an open-source conversational AI assistant built on the AI DIAL platform. It retrieves information from official Ukrstat APIs using a retrieval-augmented generation architecture and provides links to the underlying source data. The tool supports multiple languages and integrates with Ukraine’s national digital identity systems, including Diia, BankID and id.gov.ua. Released as an open-source Digital Public Good, StatGPT can be reused by other national statistical offices. Ukraine is the second institution in the world, after the IMF, to deploy an AI assistant for official statistics.

Sigma also includes an online survey module, enabling secure population surveys and providing an alternative channel for data collection when wartime conditions make traditional fieldwork difficult. API integration automates the exchange of data with administrative data suppliers, international institutions and other users of official statistics. One government ministry has already transitioned to daily automated retrieval through an API, eliminating manual file transfers.

The platform’s cloud infrastructure is hosted on Microsoft Azure outside Ukraine, with layered security measures including Zero Trust architecture, geographic redundancy and continuous monitoring. This approach strengthens resilience against risks to physical infrastructure and supports continuity during wartime.

Together, these capabilities move Ukrstat from a collection of disconnected systems towards a more integrated statistical ecosystem. Sigma is not only a technology platform, but also the digital foundation for a broader shift towards more accessible, interoperable and user-centred official statistics.

Implementing the Capacity-Building Transformation

The reform was initiated in 2021 under the Ministry of Digital Transformation, with a new law on official statistics providing the regulatory foundation. The project was delivered through the Swiss-Ukrainian EGAP Programme, administered by the East Europe Foundation. Following a competitive tender, EPAM was selected as technology partner and consultant.

The partnership brought together government ownership, international financing and technical expertise. This model provided the institutional authority and delivery capacity needed to undertake a transformation extending beyond technology into organisational design and institutional capability.

The project followed a phased approach: audit and strategy in 2021; platform development and organisational redesign, including the introduction of the SAFe methodology, in 2022; public beta launch in August 2023; and data migration and full production launch with integrated StatGPT in November 2025.

A key part of implementation was preparing Ukrstat to sustain the transformation independently. EPAM supported the creation and staffing of the Centre for Statistical Data Processing, a separate state institution designed to operate under more flexible, near-market conditions to attract specialised technical talent. The new centre has around 25 specialists, who were trained and supported during its formation.

Capability-building also extended across Ukrstat. Employees were trained in product and project management, while roles in the new statistical production process were redefined. The objective was to embed the operating practices of a modern technology organisation within a public institution, allowing Ukrstat to maintain and develop Sigma independently.

The transformation also forms part of Ukrstat’s broader strategic direction. The agency has positioned itself as an architect of the national data ecosystem, with strategic priorities that include stronger digital infrastructure, improved statistical products, more people-centred statistics and greater institutional capacity.

Tackling the Conditional and Organisational Challenges

The scale of the transition created challenges throughout implementation. One of the most significant was data migration: the complexity of transferring tens of thousands of tables was initially underestimated, affecting timelines.

Organisational change was another challenge. Staff who had worked with legacy tools and established processes needed time to adapt to the new operating model. After launch, user feedback also identified gaps in the platform’s usability. For example, the search function initially did not recognise commonly used terms such as “GDP” or “inflation” when users expected them to correspond to official statistical terminology. These issues were addressed through systematic feedback from an expert group of 35 specialists.

The pace of transformation was also shaped by the nature of official statistics. Statistical production operates according to internationally synchronised calendars and methodological requirements, creating a different rhythm from a typical technology environment.

Wartime conditions added another layer of complexity. Active hostilities disrupted field data collection and created risks to physical infrastructure. Hosting Sigma outside Ukraine and introducing alternative digital collection methods were therefore not only technology choices, but measures designed to strengthen continuity. Modern cybersecurity protections were incorporated into the new solution, while Ukrstat’s legacy information systems were affected by a major cyberattack in 2023.

New Approach to Statistics Gathering as a Measurable Impact

The transformation has established a new digital foundation for the production and use of official statistics in Ukraine:

• 2,000+ statistical indicators are available through one public portal.

• 5,000 StatGPT conversations were recorded during the first 2.5 months after launch.

• Ukraine became the second institution in the world, after the IMF, to deploy an AI assistant for official statistics.

• SDMX compliance supports interoperability with Eurostat, the IMF, the OECD and other international partners.

• A new Centre for Statistical Data Processing was established with around 25 technical specialists.

• Sigma provides the technical foundation for future reforms, including the potential use of nationwide transactional data for consumer price index calculations.

Beyond these measurable outputs, the transformation has changed the way official statistics can be produced and accessed. Data is being brought into a more integrated digital environment, while users can increasingly interact with official statistics through interfaces designed around their needs rather than the structure of legacy systems.

Equity and Inclusion of StatGPT

Making official statistics easier to access is a central element of the transformation. StatGPT’s natural-language interface reduces the specialist knowledge required to find and interpret official data, while multilingual functionality can extend access to international researchers and other users.

Integration with Ukraine’s national digital identity systems can also reduce friction in accessing digital services. At the same time, the online survey module supports continued statistical coverage across regions and population groups affected by the war by providing an alternative to traditional field-based collection.

These choices support Ukrstat’s broader objective of developing more people-centred statistics: data that is not only produced to a high standard, but is accessible, understandable and useful to the people and institutions that rely on it.

Lessons for Other Governments

Combine technical and organisational reform

Technology alone does not create institutional transformation. Redesigning roles, building new capabilities and shifting organisational culture need to happen alongside the development of new digital infrastructure.

Build institutional capacity from the outset

Knowledge transfer and the creation of internal technical capabilities can help public institutions sustain and further develop new systems after external delivery partners step back.

Use international standards as a design principle

Building around standards such as SDMX can strengthen interoperability, facilitate engagement with international partners and support alignment with wider regional and global reform agendas.

Use open source to extend public value

Releasing reusable components can allow solutions developed for one public institution to be adapted elsewhere, extending the value of public investment beyond the original use case.

Design for continuity under disruption

Cloud infrastructure, alternative data-collection channels and phased implementation can help governments maintain critical functions when conventional systems or operating environments are disrupted.

Next Steps: A Three-Year Roadmap

In 2026, Ukrstat plans to consolidate regional statistics from 25 separate websites into a single platform. The next phase also includes integration with State Tax Service scanner data to support daily price monitoring, continued migration of remaining statistical observations under a three-year roadmap and further automation of data exchange across government.

These steps are intended to build on Sigma’s role as the digital foundation of Ukraine’s statistical system and further strengthen the accessibility, interoperability and resilience of official statistics.