Modern organizations face significant challenges in managing large volumes of IT and service-related helpdesk requests. Traditional helpdesk systems rely heavily on manual categorization, delayed assignment, and slow response times, which increases operational cost and reduces user satisfaction. This paper presents an AI-Driven Smart Helpdesk System designed to automate ticket creation, classification, routing, and resolution workflow. The system integrates a REST-based ticketing API, a role-based web application, and an AI analysis engine that generates ticket summaries, detects urgency, and recommends resolution actions. A hybrid approach is implemented using both keyword-based NLP classification and large language model (LLM) reasoning using Meta Llama 3.1 via NVIDIA API. The system uses Flask, SQLAlchemy, Redis caching, and SendGridbased email notifications to ensure scalability and reliability. Experimental evaluation demonstrates improved ticket routing accuracy, reduced response time, and streamlined support operations.
Almas et al. (2026) studied this question.