This study presents an AI-driven intelligent matchmaking system that facilitates peer-assisted learning in educational communities by dynamically connecting students based on academic profiles, shared interests, and availability. The system addresses critical challenges in modern digital learning environments, including inefficient peer matching, a lack of personalization, and low student engagement. By leveraging machine learning algorithms, particularly TF-IDF vectorization and cosine similarity, the platform intelligently forms collaborative study groups that promote peer-to-peer knowledge exchange. The system integrates gamified learning features, including badges, points, and leaderboards, to sustain motivation, and incorporates real-time progress tracking through comprehensive analytics dashboards. Built using a microservice architecture with React.js frontend and FastAPI backend, the platform demonstrates superior scalability and modularity compared with traditional learning management systems. The testing results indicate the successful implementation of all core functionalities with 100% test case pass rates across the user management, matchmaking, resource sharing, and gamification modules. The system represents a significant advancement in creating student-centered, adaptive digital learning ecosystems that foster meaningful academic collaboration beyond conventional classroom boundaries.