The complex nature of the Indian legal system, with its vast array of statutes and procedural intricacies, often creates significant barriers for citizens seeking to understand their rights or navigate legal challenges. Traditional methods of preliminary legal research are manual, time-intensive, and require specialised expertise, leading to an accessibility gap between the public and legal resources. Addressing these challenges, this paper proposes an interactive legal assistance framework powered by Agentic Artificial Intelligence (AI) and Retrieval-Augmented Generation (RAG) for accessible and context-aware legal guidance. The proposed system employs a Multi-Agent Orchestration architecture to simulate a comprehensive legal workflow. Multiple autonomous agents—including a Case Intake Specialist, Advisory Guardian, IPC Section Analyst, and Legal Drafter—work collaboratively to analyse user queries, identify applicable Indian Penal Code (IPC) sections, and formulate strategic advice. An integrated RAG-based knowledge layer grounds the agents’ reasoning in actual statutes and judicial precedents, generating human-readable legal summaries and formal documentation that bridges the gap between raw facts and legal terminology. Experimental evaluation demonstrates the capability of the system to interpret plain-language descriptions and output structured, actionable legal insights with high relevance. The proposed Agentic AI framework achieves a retrieval precision of 91.3%, a hallucination rate of only 3.8%, and a legal completeness score of 9.7 out of 10, outperforming all baseline methods. The proposed framework offers a scalable, transparent, and user-friendly solution for democratising access to legal information, serving as a powerful preliminary tool for both citizens and legal professionals.

Legal Technology Agentic AI Multi-Agent Orchestration Retrieval-Augmented Generation Automated Legal Reasoning Natural Language Processing Indian Penal Code Analysis