ASSESSIQ: A UNIFIED AI-POWERED MULTI-MODAL EXAMINATION AND INTERVIEW PROCTORING SYSTEM WITH AUTOMATED EVALUATION AND STRUCTURED INTERVIEW MANAGEMENT
DOI: 
SHARE THIS ARTICLE
DOWNLOAD ARTICLE
REVIEW REPORT
CITE THIS ARTICLE
Online assessments have become integral to modern education and recruitment processes, yet existing platforms often lack flexibility, security, and intelligent evaluation capabilities. This paper presents AssessIQ, a comprehensive Flask-based web application that supports five distinct examination modalities: Multiple Choice Questions (MCQ), descriptive text-based assessments, voice interviews, webcam-based video interviews, and structured multi-section interviews. The system integrates OpenAI's Whisper model for automatic speech recognition, enabling server-side transcription of audio and video responses with high accuracy. A Large Language Model (LLM) evaluation pipeline powered by LLaMA 3.3 70B via the Groq API implements a five-criterion rubric-based scoring framework for open-ended responses, incorporating difficulty-adaptive strictness rules. Security and integrity are enforced through OTP-based email verification and a real-time proctoring module employing MediaPipe FaceMesh for six-point gaze calibration, eye-gaze deviation detection, face presence monitoring, and automated screenshot capture with timestamped warning logs. An administrative dashboard supports test creation from uploaded PDF or DOCX documents, a structured Question Bank module, multi-section structured interview management, automated AI evaluation, and result export in CSV, Excel, and PDF formats. Experimental evaluation across a cohort of 120 candidates demonstrates a Whisper WER of 6.4% for voice interviews and 7.9% for webcam-extracted audio, LLM evaluation Cohen's Kappa of 0.71 against expert human graders, and a proctoring violation detection rate of 94.8%. The platform offers a scalable, transparent, and unified solution for intelligent online assessment in both academic and professional recruitment contexts
Keywords
Article Information
Article Metrics
HOW TO CITE
Copy
S, S. M. A. (2026). Assessiq: A Unified AI-Powered Multi-Modal Examination and Interview Proctoring System with Automated Evaluation and Structured Interview Management. International Journal of Science, Strategic Management and Technology, 02(05). https://doi.org/10.55041/ijsmt.v2i5.020
S, Sheik. "Assessiq: A Unified AI-Powered Multi-Modal Examination and Interview Proctoring System with Automated Evaluation and Structured Interview Management." International Journal of Science, Strategic Management and Technology, vol. 02, no. 05, 2026, pp. . doi:https://doi.org/10.55041/ijsmt.v2i5.020.
S, Sheik. "Assessiq: A Unified AI-Powered Multi-Modal Examination and Interview Proctoring System with Automated Evaluation and Structured Interview Management." International Journal of Science, Strategic Management and Technology 02, no. 05 (2026). https://doi.org/https://doi.org/10.55041/ijsmt.v2i5.020.
References
[2] T. Nguyen, H. Le, and V. Tran, "Adaptive difficulty selection in online MCQ examination systems using item response theory," in Proc. 2021 IEEE Int. Conf. Teaching, Assessment, and Learning for Engineering (TALE), Wuhan, China, Dec. 2021, pp. 1–6.
[3] A. Kumar and R. Singh, "Automated short-answer grading using BERT-based semantic textual similarity," in Proc. 2022 Int. Conf. Advanced Computing and Communication Systems (ICACCS), Coimbatore, India, Mar. 2022, pp. 1–6.
[4] A. Radford, J. W. Kim, T. Xu, G. Brockman, C. McLeavey, and I. Sutskever, "Robust speech recognition via large-scale weak supervision," in Proc. 40th Int. Conf. Machine Learning (ICML), Honolulu, HI, USA, July 2023, pp. 28492–28518.
[5] J. Park, S. Lee, and Y. Kim, "Automatic speech recognition-based interview simulation for English oral proficiency assessment," in Proc. 2022 IEEE Int. Conf. Advanced Learning Technologies (ICALT), Bucharest, Romania, July 2022, pp. 1–5.
[6] T. B. Brown et al., "Language models are few-shot learners," in Advances in Neural Information Processing Systems (NeurIPS), vol. 33, Vancouver, Canada, Dec. 2020, pp. 1877–1901.
[7] S. Gopi, D. Sreekanth, and N. Dehbozorgi, "Enhancing engineering education through LLM-driven adaptive quiz generation: A RAG-based approach," in Proc. 2024 IEEE Frontiers in Education Conf. (FIE), Washington, DC, USA, Oct. 2024, pp. 1–6, doi: 10.1109/FIE61694.2024.10893146.
[8] J. Wei et al., "Chain-of-thought prompting elicits reasoning in large language models," in Advances in Neural Information Processing Systems (NeurIPS), vol. 35, New Orleans, LA, USA, Dec. 2022, pp. 24824–24837.
[9] H. M. Alessio, N. Malay, K. Maurer, A. J. Bailer, and B. Rubin, "Examining the effect of proctoring on online test scores," Online Learning, vol. 21, no. 1, pp. 146–161, Mar. 2017.
[10] A. Nigam, P. Pasricha, T. Singh, and P. Churi, "A systematic review on AI-based proctoring systems: Past, present and future," Education and Information Technologies, vol. 26, no. 5, pp. 6421–6445, Sept. 2021.
Ethics and Compliance
Indexed In
Similar Articles
Previous Article
