Raqman - Administrative Assistance Platform

Overview

Reduced manual form handling by 60% by designing and building an accessible full-stack administrative platform tailored for elderly and non-literate users. Accelerated request processing by 40% via automated classification, routing, and role-based access control across government departments. Implemented a RAG-based conversational assistant exposed via a FastAPI service and integrated with a NestJS backend, enabling users to submit requests via natural language conversations. Enabled fully voice-driven user journeys by integrating AI-powered OCR, speech-to-text, and document extraction pipelines to process paper-based forms and route structured data to the appropriate administrative agents.

Key Features

• Accessible Interface for Elderly & Non-Literate Users
• RAG-Based Conversational Assistant
• Voice-Driven User Journeys
• AI-Powered OCR & Document Extraction
• Speech-to-Text Integration
• Automated Request Classification & Routing
• Role-Based Access Control
• Government Department Integration

Technical Challenges

• Designing an accessible UI for non-literate users
• Integrating RAG pipeline with production backend
• Building reliable voice-driven workflows
• Processing paper-based forms with OCR accuracy
• Routing structured data across government departments