SimInterview: Transforming Business Education through Large Language Model-Based Simulated Multilingual Interview Training System

2025-08-15

AI Summary

Traditional business interview training fails to meet employers’ growing demands for personalized, cross-cultural competencies. This paper proposes a multilingual simulation interview system powered by large language models (LLMs), integrating retrieval-augmented generation (RAG) and synthetic AI agents to enable adaptive, resume–job-description–driven interviews. We introduce the novel “contestable AI” design framework—incorporating bias detection, decision explainability, and human-in-the-loop intervention—to ensure regulatory compliance and ethical robustness. The system orchestrates OpenAI o3, Llama 4 Maverick, Gemma 3, Whisper, GPT-SoVITS, Ditto, and ChromaDB to support end-to-end voice interaction and digital-human visualization. Empirical evaluation demonstrates significant improvements in interview preparedness for English and Japanese contexts, high alignment between assessment outcomes and job requirements, strong resume information retention, and high user satisfaction—where Gemma 3 achieves optimal performance under lightweight deployment constraints.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Natural Language Processing: (Large) Language Models

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Business interview preparation demands both solid theoretical grounding and refined soft skills, yet conventional classroom methods rarely deliver the individualized, culturally aware practice employers currently expect. This paper introduces SimInterview, a large language model (LLM)-based simulated multilingual interview training system designed for business professionals entering the AI-transformed labor market. Our system leverages an LLM agent and synthetic AI technologies to create realistic virtual recruiters capable of conducting personalized, real-time conversational interviews. The framework dynamically adapts interview scenarios using retrieval-augmented generation (RAG) to match individual resumes with specific job requirements across multiple languages. Built on LLMs (OpenAI o3, Llama 4 Maverick, Gemma 3), integrated with Whisper speech recognition, GPT-SoVITS voice synthesis, Ditto diffusion-based talking head generation model, and ChromaDB vector databases, our system significantly improves interview readiness across English and Japanese markets. Experiments with university-level candidates show that the system consistently aligns its assessments with job requirements, faithfully preserves resume content, and earns high satisfaction ratings, with the lightweight Gemma 3 model producing the most engaging conversations. Qualitative findings revealed that the standardized Japanese resume format improved document retrieval while diverse English resumes introduced additional variability, and they highlighted how cultural norms shape follow-up questioning strategies. Finally, we also outlined a contestable AI design that can explain, detect bias, and preserve human-in-the-loop to meet emerging regulatory expectations.

Problem

• Providing individualized multilingual interview training for business professionals
• Creating culturally aware practice matching job requirements across languages
• Overcoming limitations of conventional classroom interview preparation methods

Innovation

• LLM-based virtual recruiters for multilingual interviews
• RAG framework for dynamic resume-job matching
• Integrated multimodal AI with speech and visual synthesis