AI-Driven Dynamic Technician Ranking and Recommendation with Urgency Classification and Demand Forecasting

Traditional ranking systems for technicians rely primarily on fixed metrics such as average star ratings. These fixed metrics fail to account for the multidimensional nature of service quality as well as urgency. The goal of this paper is to make a dynamic AI-based Technician Ranking and Recommendation System (TRRS). The TRRS integrates Natural Language Processing (NLP) to perform real-time criticality triage, Machine Learning (ML) to forecast demands, and a composite scoring mechanism to evaluate technicians fairly. We utilized a DistilBERT model and fine- tuned it to classify customer requests based on urgency (High/Medium/Low) according to their textual descriptions. Additionally, we used a Random Forest Classifier to estimate demand hotspots at the village level by analysing service logs from previous services performed. The composite score for ranking technicians is composed of the quality of customer ratings, the efficiency of service duration, and the reliability of the repeatness index. We implemented the proposed solution under a microservices architecture using Node.js and Python (FastAPI). We also provide experimental results demonstrating the effectiveness of the proposed methodology through high classification accuracy (94%) for urgency criticality and (96%) for hotspot forecasting. The results prove the effectiveness of TRRS for prioritizing high-priority requests and better managing technician deployments.