AI-Driven Mock Interview: A Comprehensive Review of It’s Role in Candidate Preparation and Recruitment
Swathi Tejah Yalla, Akshitha Thokala, Akshitha Lakavath, Sai Snigdha · 2025
Recent advancements in artificial intelligence-driven systems have significantly impacted interview assessments and video streaming. Real-time emotion recognition has improved key technologies, including Convolutional Neural Networks (CNNs), natural language processing (NLP), OpenCV, and RoBERTa, contributing to enhanced video quality under varying network conditions. Notable developments include adaptive video streaming models utilizing AutoRegressive Integrated Moving Average (ARIMA) to optimize video quality in unstable environments and fairness-aware interview models designed to mitigate candidate evaluation bias. Interactive bots incorporating emotion recognition and multimodal learning offer enriched feedback, while RoBERTa enhances text similarity evaluation in interview systems. Despite progress, challenges persist, such as the accurate recognition of diverse spoken accents, biases in non-verbal communication interpretation, and data limitations, necessitating further research. This review concludes that artificial intelligence enhances accuracy, fairness, and personalization in interview assessments and video streaming.