With the growing need for more accessible legal support, this project developed an AI-powered system to recommend relevant Indian Penal Code (IPC) sections based on user complaints. We began by creating a high-quality, structured dataset, systematically cleaning and processing raw IPC information. To make our model more robust, we also generated realistic, synthetic FIRs with randomized details using advanced NLP techniques. Our system is based on a predictive model that was trained on this comprehensive dataset. Upon comparing various machine learning algorithms, we decided to use the Random Forest classifier, which had a highest accuracy of 87.25 compared to other algorithms in predicting the IPC sections. The final system takes an FIR complaint as input and identifies the most probable IPC sections with confidence scores and other legal information. Finally, this project preconditions the development of intelligent legal tools that will simplify the drafting of FIR and make legal knowledge more accessible to all people.