AI-Powered Mock Interview Web Application

Abstract

The AI-Powered Mock Interview Web Application is developed to provide an intelligent and personalized platform for interview preparation. Traditional mock interview practices are often limited by cost, accessibility, and lack of customization. The proposed system offers an on-demand web-based solution that simulates realistic interview scenarios tailored to specific job roles, skills, and experience levels. The application utilizes Google Gemini AI to generate dynamic interview questions and deliver automated feedback on user responses. The frontend is built using Next.js and React, while MYSQL manages structured data such as user profiles and interview records. Firebase is integrated for real-time services and cloud support. The system also provides performance analytics to help users track progress and identify areas for improvement. By combining generative AI with modern web technologies, the platform enhances interview readiness and provides a scalable solution suitable for educational and career development environments

Introduction

The AI-Powered Mock Interview Web Application is designed to provide an intelligent, personalized, and interactive platform for interview preparation. Traditional preparation methods such as reading questions, watching tutorials, and attending limited mock sessions often lack personalization, real-time feedback, and accessibility. The proposed system uses Generative AI to overcome these limitations by creating realistic interview simulations and providing automated performance analysis.

The application generates role-specific interview questions based on user inputs such as job role, technical skills, and experience level. It evaluates user responses using AI and provides structured feedback on strengths, weaknesses, clarity, relevance, and technical accuracy. The system aims to improve confidence, reduce interview anxiety, and enhance preparation efficiency.

The platform is developed using modern web technologies:

• Next.js and React for a responsive and interactive frontend.
• MYSQL for storing structured data such as user profiles, interview records, scores, and analytics.
• Firebase for authentication, cloud services, and real-time operations.
• Google Gemini AI as the core engine for dynamic question generation and response evaluation.

The literature review highlights previous AI interview systems that focused on facial expression analysis, speech recognition, emotion detection, and behavioral evaluation using machine learning models such as CNN and NLP. However, many existing systems rely on fixed question banks or focus mainly on recruitment evaluation. The proposed system improves adaptability by using generative AI for dynamic and personalized interview experiences.

The methodology follows a modular architecture consisting of:

• User Authentication Module: Provides secure login, profile management, and session control.
• Interview Configuration Module: Collects user details and creates customized AI prompts.
• AI Question Generation and Evaluation Module: Generates interview questions and analyzes responses using Google Gemini AI.
• Hybrid Data Management Module: Combines MYSQL and Firebase for reliable storage and scalability.
• Performance Analytics Module: Tracks user progress and displays improvement trends.

The system architecture includes five major layers:

1. Presentation Layer – Provides the user interface using React, Next.js, and Tailwind CSS.
2. Authentication Layer – Ensures secure access and user data protection.
3. Application Layer – Manages interview sessions, API communication, and system logic.
4. AI Processing Layer – Handles question generation and response evaluation.
5. Data Layer – Manages structured and cloud-based data storage.

Testing results show that the system successfully generates adaptive and non-repetitive interview questions, provides meaningful AI-based feedback, and maintains smooth performance. Compared with traditional rule-based systems, the generative AI approach provides deeper contextual analysis rather than simple keyword matching.

Conclusion

This paper presented the design and implementation of an AI-Powered Mock Interview Web Application that integrates generative AI with modern web technologies to enhance interview preparation. The system provides dynamic, role-specific question generation and automated response evalua-tion, enabling personalized and adaptive interview simulation. The layered architecture ensures secure authentication, ef-ficient backend coordination, scalable AI integration, and reliable data management through a hybrid database approach. Experimental evaluation demonstrated that the system delivers relevant questions, structured feedback, and stable performance under web deployment conditions. Overall, the proposed solution offers an intelligent, scalable, and practical platform for interview training. The architecture supports future enhancements such as voice-based interaction, advanced analytics, and expanded AI capabilities, making it suitable for academic and professional deployment.

References

[1] Y. Nag M. N., L. Chowdary K., S. L., and G. D., “AI-Driven Mock Inter-view: A New Era in Candidate Preparation,” International Journal of Advanced Research in Computer and Communication Engineering (IJAR-CCE), vol. 13, no. 11, 2024, doi: 10.17148/IJARCCE.2024.131134. [2] P. S. Khapekar, S. Bothara, T. Babar, and R. Kine, “AI Powered Mock Interview System with Real-Time Voice and Emotion Analysis,” International Journal of Novel Research and Development (IJNRD), vol. 10, no. 2, 2025. [3] “AI-Based Mock Interview System,” International Journal of Research Publication and Reviews (IJRPR), vol. 6, no. 5, pp. 4887–4893, 2025. [4] A. S. More, S. S. Mobarkar, S. S. Salunke, and R. R. Chaudhari, “Smart Interviews Using AI,” 2022. [5] D. Y. Dissanayake, V. Amalya, R. Dissanayaka, L. Lakshan, P. Sama-rasinghe, M. Nadeeshani, and P. Samarasinghe, “AI-Based Behavioral Analyser for Interviews/Viva,” 2021. [6] B. C. Lee and B. Y. Kim, “Development of an AI-Based Interview System for Remote Hiring,” 2021. [7] J. Devlin et al.,” BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding,” 2019. [8] T. Brown et al.,” Language Models are Few-Shot Learners,” 2020. [9] Google,” Gemini AI Technical Report,” 2024.

Copyright

Copyright © 2026 Hitesh Aghav, Atharva Waykar, Atharva Shinde, Atharva Shendkar, Prof. Zeba Vishwakarma. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.