PrepTalk:An AI-Powered Interview Preparation Platform for College Students

PrepTalk addresses a well-defined and practically significant gap in the interview preparation ecosystem for engineering students in India. By integrating Gemini-powered speech evaluation, rule-based filler word detection, browser-native behavioral proctoring, dynamic question retrieval via Firecrawl, WebSocket-based real-time feedback, and a persistent AI coaching interface into a single accessible platform, the system provides a preparation environment that exceeds the feature coverage of existing tools across all compared dimensions. Experimental evaluation confirms measurable improvements in filler word frequency, response clarity, and communication scores across repeated sessions. All ten system modules passed functional test cases, with real-time feedback delivered within 2.1 seconds. The platform achieves 95% module coverage of the complete placement preparation workflow — including aptitude practice, company-specific PYQs, ATS resume analysis, and AI-generated training plans — at no cost to the student. Known limitations include the Chromium browser dependency for the Web Speech API, the SQLite scalability ceiling for concurrent multi-user deployment, and the transcript-level evaluation scope that does not capture audio signal properties such as speaking pace and tonal variation. These are defined targets for the outlined future development roadmap. PrepTalk in its current form fulfills its central design objective: providing an honest, AI-powered training partner that is available on demand and delivers specific, actionable feedback calibrated to the demands of modern Indian campus placement.

[1] J. M. C. J. Joel Manuel, M. Sabi, M. Benson, G. Baburaj, and S. Saritha, "Q&AI: An AI powered mock interview bot for enhancing the performance of aspiring professionals," in Proc. Int. Conf. Recent Advances in Electrical, Electronics, Ubiquitous Communication, and Computational Intelligence (RAEEUCCI), Chennai, India, 2024, pp. 1–5. DOI: 10.1109/RAEEUCCI61380.2024.10547951 [2] "AI based mock interview system using natural language processing," IEEE Xplore, 2025. [Online]. Available: https://ieeexplore.ieee.org/document/11005032 [3] "MockMate: Leveraging AI for adaptive and realistic interview practice," IEEE Xplore, 2025. [Online]. Available: https://ieeexplore.ieee.org/document/11382795 [4] "AI powered virtual job interview simulator using natural language processing," IEEE Xplore, 2025. [Online]. Available: https://ieeexplore.ieee.org/document/11013504 [5] R. Mandal, P. Lohar, D. Patil, A. Patil, and S. Wagh, "AI-based mock interview evaluator: An emotion and confidence classifier model," in Proc. Int. Conf. Intelligent Systems for Communication, IoT and Security (ICISCoIS), Coimbatore, India, 2023, pp. 521–526. DOI: 10.1109/ICISCoIS56541.2023.10100589 [6] P. K. Mishra, A. K. Arulappan, I.-H. Ra, T. M. L., G. R. G., and Y.-S. Lee, "AI-driven virtual mock interview development," in Proc. Joint 13th Int. Conf. Soft Computing and Intelligent Systems and 25th Int. Symp. Advanced Intelligent Systems (SCIS&ISIS), Himeji, Japan, 2024, pp. 1–4. DOI: 10.1109/SCISISIS61014.2024.10760210 [7] B. Jadhav, A. Sawant, A. Shah, P. Vemula, A. Waikar, and S. Yadav, "A comprehensive study and implementation of the mock interview simulator with AI and pose-based interaction," in Proc. 1st Int. Conf. Cognitive, Green and Ubiquitous Computing (IC-CGU), Bhubaneswar, India, 2024, pp. 1–5. DOI: 10.1109/IC-CGU58078.2024.10530717 [8] E. M. AlGhamdi, Y. Li, D. Gasevic, and G. Chen, "Leveraging prompt-based LLMs for scoring and feedback in educational contexts," ScienceDirect, 2025. [9] FastAPI Documentation, "FastAPI: Modern, fast web framework for building APIs with Python 3.6+," 2024. [Online]. Available: https://fastapi.tiangolo.com/ [10] Google AI, "Gemini API Documentation — google.genai Python SDK," 2025. [Online]. Available: https://ai.google.dev/ [11] Firecrawl, "Firecrawl developer documentation — web scraping and extraction API," 2024. [Online]. Available: https://docs.firecrawl.dev/ [12] Mozilla Developer Network, "Web Speech API — SpeechRecognition Interface," 2024. [Online]. Available: https://developer.mozilla.org/en-US/docs/Web/API/SpeechRecognition