The faculty training programs offered by the traditional systems are generic and do not address the personalized needs of digital competencies of the teachers. The proposed research is a prediction of the AI-based system which is likely to identify the effectiveness of faculty training ranking through training records on a survey of more than 4000 faculty training records acquired in Kaggle Dataset basing on the existence of trained faculty or not, rated as either exceeds, fully met, needs improvement, and PIP which denotes Personnel Improvement Program. The model was trained using Random Forest and had an overall accuracy of 78% and weighted F1-score of 0.69 and the best predictive variables were Training Cost, Trainer and Job Function. These findings draw attention to the necessity of considering the problem of the lack of a balance between classes and the potentiality of AI in promoting individualized cost-effective training recommendations. The results form the basis of implementation of adaptive faculty development systems that maximize the available institutional resources as well as improving performance in teaching.