Victor-Valentin ANGHEL1, Theodor BORANGIU2, Cătălin NEGULESCU3
Abstract. This paper describes a software system that supports recruitment in the IT domain, where the number of applications is usually high and resumes come in many different formats. Because of this, going through them manually becomes difficult and, in many cases, inefficient. In practice, recruiters often need to process large amounts of information in a short time, so some level of automation becomes necessary.The proposed solution has two main steps. First, resumes are processed and assigned to common IT roles such as Data Scientist, Machine Learning Engineer, DevOps Engineer, Software Developer, QA Engineer, Project Manager or Cloud Architect. For this, an XGBoost model is used that considers several types of information extracted from resumes, like education, skills, programming languages, certifications, foreign languages, and previous experience. These information are combined to get an estimate of how well a candidate might fit a certain role. A scoring step is then added where the recruiter can adjust the importance of each feature. This is done because it was noticed that, in real hiring scenarios, not all criteria matter equally and their relevance can change from one position to another. The service system, currently under development, operates in a containerized environment and is orchestrated through Kubernetes technology to ensure scalability and component isolation.
Keywords: Kubernetes, Natural Language Processing (NLP), Named Entity Recognition (NER), Automated Recruitment, AI Model Fine-Tuning.
DOI 10.56082/annalsarsciinfo.2026.1.50
1 Ph.D. student, POLITEHNICA Bucharest, victor.anghel@stud.acs.upb.ro
2 University Professor, Ph.D., POLITEHNICA Bucharest, Academy of Romanian Scientists, theodor.borangiu@upb.ro
3 Ph.D. student, POLITEHNICA Bucharest, catalin.negulescu@stud.electro.upb.ro
PUBLISHED in Annals of the Academy of Romanian Scientists Series on Science and Technology of Information, Volume 19, No1