Big Data for Labour Market Intelligence in Tunisia — Machine-Classified Job Vacancies for VET Planning
Tunisia
The EU's European Training Foundation scrapes Tunisian online job ads and classifies them with machine learning into ISCO, ESCO, NACE …
Finland · Helsinki · See the Finland profile · See the Helsinki profile
Top 73% 40/100 · Ask Evidence Copilot about this practice
Finland's continuous-learning agency runs a public AI service that mines job advertisements to map skill demand, skill clusters and occupational transitions, so VET and higher-education providers can reshape what they teach. Beta launched December 2023; no impact data yet.
Osaamistarvekompassi (Skills Needs Compass) is a public, AI-assisted skills-anticipation service operated by Jatkuvan oppimisen ja työllisyyden palvelukeskus (Jotpa), the Finnish Service Centre for Continuous Learning and Employment. Jotpa is a state agency steered jointly by the Ministry of Education and Culture and the Ministry of Economic Affairs and Employment, operating in connection with the Finnish National Agency for Education. The service mines job advertisements and other large datasets to produce an annually updated picture of labour-market skill demand, skill clusters and occupational transitions. Its stated audiences are education providers in vocational and higher education, who are meant to use it to reshape provision, and career information, guidance and counselling services. It also informs how Jotpa targets its short-course training funding. The project ran from 2022, launched in beta in December 2023, and was built out in phases to summer 2025. Vendors are Solita Oy for the data platform, Digitalist Oy for the service, and Headai Oy for the AI and natural-language skills extraction.
The published figures are modest and mostly descriptive. Jotpa's April 2024 release reports 28,000 skills in the underlying skills database. Development was funded through the EU Recovery and Resilience Facility under NextGenerationEU across 2022 to 2025, but the budget amount is not published on Jotpa's pages. User and visitor numbers are not published. The classification backbone is the Finnish National Agency for Education's 2022 skills classification, itself constructed from ESCO, O*NET and the Austrian AMS Berufsinformationssystem together with job-advertisement data.
Evidence of effect is essentially absent. There is no impact evaluation, no controlled study, and no published take-up statistics. The OECD Observatory of Public Sector Innovation's case entry from June 2024 states plainly that the beta version had been in use for only a month and there were no results yet. The only evaluative work identified is a commissioned user-experience study run by LabWellTech using eye-tracking, think-aloud protocols and service-design interviews, with participant numbers not disclosed; its finding was that the interface and content were too complex for the professionals the service targets, and improvements shipped in autumn 2025.
Jotpa's own material is candid about the method's limits: incompatible classifications between ESCO and national systems, the quality assurance burden of noisy big data, and survey bias. Job-advertisement data structurally under-represent sectors and roles that do not recruit online. A further open question is sustainability, since development was Recovery and Resilience Facility funded to 2025 and continuation depends on the national budget. The value of the case is as a governance and design model — a state agency rather than a vendor owning the skills-anticipation data used to steer public education provision, with limitations disclosed rather than hidden — not as demonstrated impact on curricula or learners.
Read the full analysis: https://www.osaamistarvekompassi.fi/
Implementation detail (cost, timeline, staffing, conditions for success) is not yet available for this practice.
Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.
Where this practice's information was retrieved from, and when.
Tunisia
The EU's European Training Foundation scrapes Tunisian online job ads and classifies them with machine learning into ISCO, ESCO, NACE …
Indonesia
A 12-week trial with 180 Indonesian vocational students (90 intervention, 90 control) found an AI system combining skills-mapping, chatbot mentoring …
Peru
The World Bank, Peru's DRELM, Anthropic and Microsoft launched an AI maths tutor and career-guidance coach across 110 Lima public …
Nigeria
KLCI's Rafiki AI, a free WhatsApp generative-AI career advisor on Amazon Bedrock, supports underserved and displaced Nigerian and African youth …
Open full copilot Grounded in cited practices — always check the sources.