DFROG — De Nederlandsche Bank's Machine-Learning-Tested Nowcasting Model for GDP Growth
Netherlands
DNB's DFROG nowcasting model has tracked Dutch GDP growth in real time since 2024, publishing monthly estimates; a decade-long evaluation …
Norway · Oslo · See the Norway profile
Since 2019 Norway's Storting has built StorSak, a digital value-chain replacing paper case handling; in 2025 it began piloting a machine-learning module that auto-tags parliamentary cases and documents by subject to decide what must be archived.
In 2019, the administration of Norway's Storting (parliament) adopted a strategic digitalisation goal: to exchange, process, publish and preserve parliamentary information digitally rather than on paper. This gave rise to two linked programmes — SIDA, covering the administration's internal document culture, and StorSak, which targets the parliamentary case-handling process itself, from a case's receipt through committee review, plenary debate, decision and dispatch.
One investment area within StorSak is the use of machine learning for what the project calls 'theme-setting' (temasetting): automatically classifying cases and documents by subject and using predetermined criteria to decide which steps of a case must be preserved as files or metadata in the parliament's archive. Described by Stortinget's own archival staff as the chamber's 'first attempt at using artificial intelligence in archive creation,' the AI module was scheduled to begin piloting in autumn 2025.
The initiative is documented in a peer-reviewed case study in the International Journal of Parliamentary Studies (2024) and in Norwegian professional archival-management literature, both of which describe StorSak as an ongoing, multi-year digitalisation effort rather than a finished, evaluated product. As of the most recent published account, no accuracy, adoption or efficiency figures for the AI theme-setting component had been released, and its use remained confined to a pilot phase — a limitation this entry flags explicitly.
Read the full analysis: https://brill.com/abstract/journals/parl/4/1/article-p79_005.xml
Implementation detail (cost, timeline, staffing, conditions for success) is not yet available for this practice.
Where this practice's information was retrieved from, and when.
Netherlands
DNB's DFROG nowcasting model has tracked Dutch GDP growth in real time since 2024, publishing monthly estimates; a decade-long evaluation …
Poland
Narodowy Bank Polski scrapes millions of online retail prices to nowcast food inflation ahead of official CPI releases. Peer-reviewed tests …
Croatia
Facing reform fatigue from repeated manual surveys, Croatia's Ministry of Economy and the World Bank piloted an AI/NLP algorithm scanning …
Canada
Bank of Canada researchers used ML on payments-system data to nowcast GDP and trade during COVID-19, cutting forecast error 20–40% …
Open full copilot Grounded in cited practices — always check the sources.