AI-Supported Interview Training for Young People with Disabilities: The ELVIR Virtual Trainer
Abstract
Young people with disabilities often face barriers when preparing for job interviews, particularly due to limited opportunities for repeated practice and individualized guidance. Recent advances in generative artificial intelligence and large language models offer new possibilities for creating conversational training environments that complement traditional employability programs. This paper presents ELVIR (Virtual Job Interview Trainer), an artificial intelligence-based platform designed to support interview preparation for young people participating in employment inclusion programs. The system was developed in collaboration with specialists from Teletón Chile and integrates conversational AI, virtual avatars, configurable interview scenarios, and session recording capabilities to provide simulated job interview experiences. Users can select different occupational profiles and interview difficulty levels, allowing the platform to support a variety of training needs and preparation contexts. The contribution of this work is the design and implementation of a virtual interview training platform specifically oriented toward employment inclusion. The paper describes the motivation for the system, its architecture, implementation, and current deployment within a pilot program conducted with participants from Teletón Chile. In addition, the paper discusses the opportunities and challenges associated with the use of conversational AI to support interview training in disability inclusion contexts. Keywords: disability inclusion, employment inclusion, artificial intelligence, conversational agents, job interview training