Career Recommendation System (CRS) is an artificial intelligence solution capable of suggesting appropriate jobs or careers based on user profiles and industry needs. This study presents a systematic literature review that focused on variant characteristics of CRS and has been implemented in the last ten years. The review found 17 studies were extracted from ACM, IEEExplore, Science Direct, Springer, Willey, and MDPI databases. The results of this review prove that a hybrid recommender system is the most frequently (47%) approach implemented in CRS studies. Text mining (29,5%) is most commonly applied as the artificial intelligence technique in CRS. At least 7 features are needed to build a CRS model, but the most widely used are job profiles and course profiles with 71,42% and 35,71% frequency respectively. The most widely applied evaluation metrics is precision (21%), followed by acceptability, accuracy, and user response each 14% in review.