OralIQ: AI-Based Oral Examination and Interview Assessment System

OralIQ is an AI-powered platform designed to automate oral examinations and mock interviews through intelligent, objective, and real-time assessment. The system leverages state-of-the-art Artificial Intelligence technologies to evaluate a student's performance from multiple perspectives, reducing human bias and improving the fairness and consistency of oral assessments. The platform supports two primary user roles: Lecturers and Students. Lecturers can create exams, add questions, define model answers, monitor student participation, and review detailed AI-generated analytics. Students can join oral exams through a secure invitation link, answer questions using their microphone and webcam, and receive personalized feedback along with a detailed performance report after completing the exam. OralIQ employs a multimodal AI pipeline consisting of three main components. OpenAI Whisper converts spoken responses into text while analyzing speech characteristics such as fluency, confidence, and speaking rate. Sentence-Transformers (all-MiniLM-L6-v2) evaluates the semantic similarity between the student's response and the lecturer's model answer using sentence embeddings and cosine similarity. Meanwhile, MediaPipe Face Mesh, OpenCV, and DeepFace monitor facial expressions, gaze direction, head pose, and emotional state to assess attention and detect potential suspicious behavior during the examination. The final grade is calculated using a weighted evaluation strategy: Content Analysis – 40% Speech Analysis – 30% Facial Expression Analysis – 30% The system is built using React.js (Vite) for the frontend, FastAPI for the backend, SQLite for development, PostgreSQL for production, and is deployed using Vercel and Hugging Face Spaces. By combining speech recognition, natural language processing, and computer vision, OralIQ provides an intelligent, scalable, and explainable solution for modern educational assessment and AI-powered mock interviews.