Name in original language
ភ្នាក់ងារបញ្ញាសិប្បនិម្មិតនៃថ្នាលវិភាគទិន្នន័យសេដ្ឋកិច្ចនិងហិរញ្ញវត្ថុ
Initiative overview
The Data EF Open Data Portal serves as Cambodia’s central platform for sharing economic and financial datasets across areas such as public finance, macroeconomics, revenue collection and industry. While the portal provides extensive data access, staff and decision‑makers often struggled to translate these datasets into timely insights. Manual search, review and download processes were slow and resource‑intensive, and inconsistent metadata and complex schemas made it difficult for users to locate relevant information. These limitations reduced the portal’s effectiveness as a decision-support tool.
To address this gap, the General Department of Digital Economy developed the Data EF AI Agent and integrated it into Data EF Insight, the analytics portal. The AI agent connects natural-language queries to underlying datasets through a semantic metadata layer, automatically identifying relevant data, retrieving it from the Data EF lakehouse and delivering results as tables, visualizations or concise summaries. This allows users to move directly from a question to an insight within seconds, while keeping human oversight at the center of validation and decision‑making.
The Data EF AI Agent aligns closely with Cambodia’s digital transformation agenda. It supports the Digital Government Policy 2022–2035 and the Digital Economy and Society Policy Framework 2021–2035, both of which emphasize evidence-based policymaking, data interoperability and digital government. Cambodia’s low scores on the OECD OURdata Index (0.04) and the Digital Government Index (0.2) highlight the importance of initiatives that make government data easier to access and analyze. The AI agent reduces barriers to data use, strengthens analytical capacity and advances responsible, human-centered AI consistent with OECD AI Principles.
The purpose of the AI agent is to remove technical barriers and improve administrative efficiency across the Ministry of Economy and Finance (MEF). Internal testing shows that data retrieval and summarization can now be completed in 10–15 seconds—compared to lengthy manual processes involving multiple dataset searches and downloads. Staff can explore more than 600 datasets through a single, easy‑to‑use interface, strengthening evidence-based reporting and analysis. Improvements are evaluated through workflow comparisons, usage observations and user feedback on usability and efficiency. Early results show reduced reliance on technical expertise, faster reporting cycles and stronger internal data-use capability. The initiative is expected to encourage broader adoption of data-driven governance within MEF and expand to additional use cases as confidence grows.
