The rapid advancement of large language models (LLMs) offers transformative potential for personalized education, yet existing e-learning platforms for coding and placement training remain largely static and generic. This paper presents an AI-integrated platform that leverages the Google Gemini API to deliver adaptive coding challenges with intelligent hints, automated code evaluation, realistic mock interviews with dynamic feedback, resume parsing and job matching, and personalized course recommendations. Built on a modern full‑stack architecture (React, Node.js, MongoDB, Google OAuth), the system maintains a longitudinal learner profile that tracks progress across all modules. A pilot study with 30 final‑year computer science students over four weeks demonstrated significant improvements: coding scores increased by 16.4 points, interview scores rose by 2.4 points, and resume ATS scores improved by 23%. The paper details prompt engineering strategies, system architecture, implementation challenges (hallucination, over-reliance, privacy), and mitigation techniques. This work provides a replicable blueprint for integrating state‑of-the-art LLMs into vocational training, transforming one-size-fits-all preparation into adaptive, learner-centric skill development.