Design and implementation of an AI-powered legal assistantplatform using retrieval-augmented generation and agenticworkflows

Abstract

Access to accurate legal information and finding verified legal representation can be challenging for individuals navigating complex legal documents and procedures. Conventional legal research often involves parsing dense statutory text and case law manually, while finding nearby legal professionals lacks integrated contextual support within digital research tools. This paper presents an integrated legal technology platform designed to streamline legal research and client-lawyer connectivity through Retrieval-Augmented Generation (RAG), intelligent agent workflows, and a geolocalized discovery system. The proposed system utilizes direct document processing and vector-based semantic retrieval to analyze legal PDF texts and provide precise, context-aware answers via an LLM-powered agent orchestrator. In addition, the platform features a location-based lawyer discovery module using Haversine distance calculations to help users locate nearby legal counsel, complemented by a secure direct messaging infrastructure for client-lawyer communication. A prototype implementation demonstrates the functional integration of these components into a unified web application. Future work will focus on evaluating retrieval precision, response latency, and system scalability under real-world usage scenarios.