UniPredict: A Multi-Algorithm AI Framework for Admission and Program Prediction Across Leading Institutes of Pakistan , Trained on Real, GenAI & LLM-Synth Data Assessed using Advanced Evaluation Metrics like MCC, Cohen's kappa, Jaccord Score and Conf

UniPredict: A Multi-Algorithm AI Framework for Admission and Program Prediction Across Leading Institutes of Pakistan , Trained on Real, GenAI & LLM-Synth Data Assessed using Advanced Evaluation Metrics like MCC, Cohen's kappa, Jaccord Score and Confusion Matrix. The university admission process in Pakistan is an overwhelming challenge, plagued by uncertainty, misinformation, and a lack of standardized criteria. Students are forced to rely on guesswork, outdated merit lists, and peer opinions, often leading to wasted opportunities, financial strain, and misguided career choices. Many apply to universities where they have little chance of acceptance, while others unknowingly miss better opportunities. The absence of data-driven guidance also leaves students struggling to choose the right academic field, often leading to dissatisfaction and career mismatches. This broken system fosters confusion, stress, and inefficiencies, making the need for an AI-driven predictive solution more critical than ever. To address this, we propose UniPredict, an AI-powered system that predicts admission and recommends academic fields across 8 top universities in Pakistan. Using 16 machine learning models—Logistic Regression for admission prediction and Multi-Class Classification for field selection—across 80+ programs with high accuracy. The system is deployed as a web application, built with Next.js, Tailwind CSS, and Framer Motion, offering a seamless and user-friendly experience. Features include form validation, real-time predictions etc. Our results show that UniPredict significantly improves decision-making by providing reliable, AI-driven insights, reducing uncertainty in university admissions. Future improvements include expanding to more universities and integrating with official admission portals for a fully automated system.