The JobKG project aims to comprehensively analyze the Romanian labor market using data available on online recruiting platforms deploying state-of-art approaches in natural language processing (NLP), semantic web, and agent-based modeling (ABM). For this purpose, it performs an extensive quantitative and qualitative analysis of the Romanian labor market to prospect the opinion of various groups of stakeholders regarding the alignment and calibration of skills demanded by the labor market and those developed and trained in the education system. Given that employment-oriented online services are the main source of information for job seekers willing to search and apply for vacant positions, they represent a rich data source for understanding the occupations in demand and the relevant skills required. The project will create a knowledge graph of the Romanian labor market, starting from existing taxonomies and extracting data using advanced NLP techniques. An ABM approach will analyze the labor market dynamics, considering various scenarios and defining agents with characteristics similar to those of the entities (job seekers and firms), which will simulate the demand and supply.