Over the past three years, amid an ongoing conflict that brings weekly waves of missile and drone attacks against its cities, Ukraine has been quietly implementing a massive and complex Public Investment Management (PIM) reform. While executing a structural overhaul of this magnitude would be a daunting challenge for any nation in peacetime, Ukraine is doing so while facing reconstruction costs estimated by international bodies to exceed US$ 588 billion. Thousands of individual projects—spanning local communities, regional authorities, national ministries, and numerous international partners—are currently being planned, coordinated, and executed simultaneously across the country.
At the very center of this ambitious reform effort is DREAM, Ukraine’s open digital ecosystem for public investment management. Built squarely on the principles of open government and open contracting, the DREAM platform serves as a vital connective tissue, linking state-owned and external registers, facilitating project monitoring, and supporting rigorous data analysis. In essence, it has created a single digital pipeline for all public investment projects nationwide. Today, DREAM operates as the digital backbone of the entire reform, supporting more than 12,900 public investment projects distributed across 18 distinct sectors. The platform is actively utilized by more than 1,300 municipalities, every single regional administration, all national ministries, and major state-owned enterprises.
The platform is designed to cover the full lifecycle of public investments: guiding initiatives from initial project conception and feasibility appraisal through to strategic prioritization, funding coordination, and ongoing implementation monitoring. As this foundational reform transitions into its full-scale implementation phase, Ukrainian officials and international partners are now exploring how artificial intelligence can serve as its next practical enabler. The goal is to strengthen institutional capacity, make complex PIM rules accessible and usable for every municipality in the country, and drastically improve the overall quality, planning, and delivery efficiency of investment projects.
The Scale of the Challenge
This unprecedented scale naturally brings significant practical capacity challenges. Public officials on the ground must constantly navigate an intricate maze of complex laws, regulatory methodologies, public procurement rules, construction norms, financial requirements, and strict international donor conditions.
Local governments, particularly outside major urban centers, often operate with limited staff, fragmented technical expertise, and very little time to parse hundreds of pages of dense legal texts, methodologies, and operational manuals. For instance, completing a full cycle of an investment pre-feasibility study for a relatively modest local project typically demands about four weeks of full-time work for a single local government team. Consequently, preparing just ten separate projects becomes an overwhelming hurdle for a small municipal administration.
During peak planning and submission cycles, the central DREAM Project Office handles more than 850 support requests per day, even with extensive training programs, dedicated helpdesks, and expert networks already in place. To put the educational outreach into perspective, during 2025 alone, organizers held 96 training events reaching nearly 30,000 participants, alongside continuous helpdesk support. Traditional support models are simply struggling to scale at the rapid pace that this nationwide reconstruction requires.
This persistent capacity shortage is made acutely worse by wartime pressures. More than 230 local communities—representing roughly 15 percent of the country—currently hold front-line status, facing dense shelling, ongoing drone attacks, severe energy infrastructure breakdowns, and continuous waves of internally displaced persons arriving daily. Crucially, all of these frontline communities must still be fully included in the PIM process and retain equitable access to vital recovery financing.
When administrative capacity is stretched this thin, even the most soundly designed rules and methodologies become exceptionally difficult to apply in practice.
Our Approach to AI
Artificial intelligence cannot replace public servants, human expert judgment, or political accountability, at least not in the foreseeable future. However, it can dramatically improve how critical knowledge—such as policies, methodologies, regulatory requirements, and sectoral expertise—is accessed, interpreted, and applied, particularly in a fast-evolving and highly complex reform environment. Closing this critical gap is the primary objective of the new initiative.
Within the DREAM ecosystem, artificial intelligence is not viewed as an external automation novelty, but rather as a foundational support layer embedded directly into daily governance workflows. Its ultimate purpose is straightforward to explain yet difficult to achieve: to help local users prepare and appraise better projects, understand regulatory requirements much faster, and make more consistent, evidence-based decisions.
Three years of intense reform implementation have provided the project team with something that most nascent AI initiatives lack entirely: a solid, battle-tested foundation. The system benefits from a verified knowledge base, a clear mapping of where users routinely encounter bottlenecks, and the hard-won operational data necessary to build an assistant that is genuinely useful. Rather than serving as a superficial feature meant to attract attention, the AI tool functions as a practical mechanism to support and empower the people tasked with rebuilding the nation.
Backed by strong political support from the Ministry for Development and the Ministry of Economy, the DREAM Project Office has joined forces with the Open Contracting Partnership—drawing on its recognized expertise in deploying AI tools for public procurement—alongside uStudio as the IT and AI innovation partner, and GCIEP and PwC, with support from the UK’s Foreign, Commonwealth & Development Office (FCDO). Together, these organizations are developing an advanced AI-powered assistant fully integrated into the DREAM ecosystem.
The overall approach is grounded in a working prototype, a carefully curated and verified knowledge base, and a hybrid technical architecture that strictly aligns with Ukraine’s data sovereignty standards and the ethical AI principles outlined in the official White Paper on Artificial Intelligence Regulation in Ukraine.
The DREAM AI Assistant: Three Stages of Implementation
The rollout of the DREAM AI Assistant is strategically aligned with the broader PIM Reform Roadmap spanning from 2024 to 2028, advancing through three distinct and deliberate stages of implementation.
In 2026, the AI assistant will deliver real-time guidance regarding PIM rules and the project preparation process directly within the DREAM interface. During this initial phase, the primary objective is to bridge the gap between complex legal language and practical municipal application by translating dense methodological requirements into clear, accessible explanations tailored precisely to each user’s specific role, operational sector, and geographic location. Serving as a trusted, round-the-clock entry point into the comprehensive PIM knowledge base, the assistant provides immediate, practical guidance for government bodies at every administrative level.
Initial testing of the DREAM AI Assistant took place in April 2026. Focus group participants representing various local communities tested the assistant during actual project preparation workflows, awarding the tool an average satisfaction score of 4.33 out of 5 points. While ongoing tailoring and technical upgrades continue, it has already become clear that the assistant successfully reduces administrative workload and accelerates procedures by guiding DREAM users seamlessly through project preparation.
Looking ahead to 2027, the DREAM AI assistant is scheduled to expand into quality assurance functions. In this second stage, it will assist evaluation and appraisal teams in identifying missing indicators or internal inconsistencies within feasibility studies, while proactively suggesting ways to better align proposed projects with national sectoral priorities. Furthermore, it will help improve the economic, environmental, and financial dimensions of submissions by comparing new project proposals with historical, validated cases, thereby enhancing the overall consistency and transparency of thousands of investment decisions.
In the final targeted phase spanning 2028 through 2029, the initiative aims to introduce advanced investment modeling capabilities. Utilizing geographic, demographic, fiscal, and other socioeconomic data, the system will help public officials explore diverse development scenarios and model alternative investment options. This advanced functionality will allow authorities to conduct scenario-based assessments to determine precisely where public resources can achieve the greatest possible societal impact.
Building a Safe and Responsible AI Architecture
Because public investment decisions involve high stakes, substantial financial resources, and sensitive public data, the DREAM AI assistant is built on a hybrid, fully auditable architecture designed to ensure responsible and secure governance.
The core of the system relies on Retrieval-Augmented Generation (RAG), which ensures that every response generated is grounded exclusively in verified laws, official manuals, and approved guidelines. To safeguard sensitive information, the system employs localized language models hosted on secure infrastructure, thereby protecting national data sovereignty. Every answer provided by the system is fully traceable to an approved source, timestamped, and subject to strict version control. In addition, the overarching architecture incorporates a human-in-the-loop validation process to guarantee that all content remains accurate and reflects the absolute most recent regulatory updates.
To minimize the inherent risks of AI hallucinations or the accidental use of outdated administrative guidance, responses are systematically linked to approved regulatory sources. Potential operational risks, such as factual inaccuracies or privacy concerns, are proactively mitigated through structured prompt engineering and rigorous access controls.
Furthermore, a dedicated multidisciplinary oversight team continuously monitors the system’s performance to ensure strict adherence to ethical standards and to maintain its core focus on public service and accountability.
Ukraine’s Lesson for the World
Ukraine’s ongoing experience demonstrates that effective implementation of artificial intelligence in government does not begin with sophisticated algorithms. Instead, it starts with clear rules, robust institutions, and the operational discipline required to design technology that directly serves a public purpose. The AI layer currently being integrated into DREAM is only as effective as the underlying reform infrastructure beneath it—an infrastructure that required three years of intense labor and institutional commitment to construct.
If executed successfully, this initiative has the potential to become not only a vital catalyst for Ukraine’s national recovery and long-term development, but also a globally relevant test case for how artificial intelligence can responsibly strengthen public investment management systems in complex, high-pressure environments. A nation rebuilding under fire, while simultaneously maintaining a cutting-edge digital governance framework, proves precisely what is possible when technology meets unwavering public resolve.
