The ILO PROSPECTS Egypt Skills Profiling Application is an AI-driven career guidance tool developed by SkillLab, a social enterprise based in Amsterdam, the Netherlands, and deployed in Egypt through a partnership between the International Labour Organization (ILO) PROSPECTS programme, SkillLab, and the Association of Businessmen in Alexandria (ABA). The application was piloted between 2020 and 2021 to assist refugees and host community members in Egypt in identifying, articulating, and communicating their employable skills to potential employers, with the broader goal of facilitating labour market integration for displaced and disadvantaged populations.
The core AI functionality of the application centres on an automated skills interview powered by natural language processing (NLP), graph theory, and Bayesian machine learning. The system is built on the European Skills, Competences, Qualifications and Occupations (ESCO) classification framework, which contains approximately 13,485 skills mapped to 2,942 occupations. During the profiling process, users engage with an AI-driven conversational interface that guides them through a structured self-assessment of their work history, education, and competencies. The underlying machine learning pipeline uses NLP and graph theory to exploit intrinsic relations within the ESCO skill dataset, a hybrid collaborative recommender system to intelligently query the most promising skills from the full ESCO taxonomy, and Bayesian machine learning algorithms that adapt to each user's input as the interview progresses. These Bayesian optimisation algorithms balance between exploiting the existing ESCO dataset and exploring possible new trends not captured in existing data. All machine learning components run on Google Cloud Platform and are written in Python.
At the conclusion of the profiling process, the application generates three key outputs for each user: a comprehensive list of their identified skills, a list of potential occupations matching those skills, and a professionally formatted CV that faithfully reflects the user's indicated skills, experience, and education. The system also assesses role fit by comparing individual skill profiles against occupation requirements, and identifies training gaps where additional skills development would enable access to further occupational pathways. Employment counsellors and career guidance staff use these outputs to inform their advisory sessions with job seekers, making the system an advisory and decision-support tool rather than one that makes autonomous placement or eligibility decisions.
The pilot was conducted among approximately 400 beneficiaries in the governorate of Alexandria, Egypt. ILO PROSPECTS Egypt subsequently integrated the application into Job Search Clubs (JSCs) operated in partnership with Caritas, establishing 16 JSCs between August 2021 and January 2022 in Alexandria and Damietta for asylum seekers and refugees. In the JSC deployment, 303 participants successfully used the skill profiling tool and extracted professional CVs.
A tracer study conducted by the ILO three months after the initial pilot (published in 2022) collected direct feedback from the 400 pilot beneficiaries. The study found that 64 percent of participants used the application to create a CV, 61 percent found it very easy or easy to use, and results on perceived usefulness for job search and training identification were mixed. In the subsequent JSC deployment, 76 percent of surveyed participants rated the application as good or excellent in assisting them to identify their skills, more than 75 percent found it good or excellent for CV writing, 64 percent discovered new potential employment fields matching their skills, and 78 percent found the application user-friendly. Four months after the JSC project concluded, 88 participants had found employment in occupations including marketing, sales, nursery teaching, electrician, tailor, and clinic assistant.
The human oversight model is explicitly human-in-the-loop (HITL). The application outputs serve as advisory inputs for employment counsellors, who retain professional judgement over career guidance recommendations. No automated entitlement, benefit, or placement decisions are documented. The ILO tracer study recommends localisation of terminology, induction training for new users, employer outreach to increase awareness of the skills profiles, and longer-term evaluation before wider roll-out.
SkillLab won the Google AI Impact Challenge in 2019, receiving financial support to further develop its AI-based skills assessment technology. The platform was recognised in the IRCAI Global Top 100 AI for Sustainable Development Goals index. The implementing model involves SkillLab as the third-party technology provider, with the ILO PROSPECTS programme funding the deployment and ABA handling local pilot delivery. The application targets refugees, migrants, mature workers, informal workers, and youth not in employment, education, or training (NEET).