Job Interview Simulator
Developing AI -Powered Agents for Realistic Interview Training
YUSHIEN, Rutooja Khamkar, Jiayun Qiu, 2024
Abstract
The development of artificial intelligence (AI) has made human-agent interaction (HAI) more diverse than ever before. Chatbots, robots, and virtual agents have transformed the way humans interact with the systems. Today, AI often appears as an "agent" with purpose, motivation, and intention. A recent example of the influence of AI is the release of OpenAI's ChatGPT language model. It is remarkable that it can produce responses that are like human language. It has been applied in a wide range of areas for a lot of different work, such as customer service, inventory optimization, data analytics, recommendation systems, and supply chain management.
This paper focuses on creating AI powered virtual agents that can simulate real-life job interviews. By deploying game engine and AI technology, we wish to provide users with realistic and interactive training. It aims to simulate human-like responses and behavior, offering a realistic and dynamic environment for interview practice. The system's design incorporates natural language processing, machine learning algorithms, and advanced animation techniques to create a virtual interviewer that can pose relevant questions, respond to user’s inputs, and provide feedback based on user’s background. The goal of this project is to enhance the users' interview skills, reduce anxiety, and prepare them for successful real-life job interviews. Finally, this paper also dives into the impact of AI agent's appearance, the potential for interaction between AI agents, and the possibilities and limitations of this technology.
Keywords: Believable AI agent, conversational AI, human-agent interaction, generative AI, AI NPC, large language models.
Click to read the paper