Italiano Development of the AI-driven platform for the integrated management of training evaluation INSIGHT

Indicators and Stakeholders Integration for Generative Evaluation and Holistic Training

Abstract

The paper analyzes the main criticalities in traditional training evaluation methods and proposes an innovative model based on Artificial Intelligence (AI). Conventional evaluation is often fragmented, generic, and misaligned with performance outcomes, while established models like those of Kirkpatrick and Phillips have significant limitations in their practical application. The following analysis identifies methodological, technological, and cultural barriers, proposing a system that redefines evaluation as a strategic and integrated process. The proposed model employs conversational AI to gather expectations, monitor the evolution of learning, and suggest improvements through a combination of objective and subjective indicators. Its uniqueness also lies in the AI-human supervision hybrid approach, multi-stakeholder validation, and the use of natural dialogue, aiming to transform evaluation into a tool for managing and developing organizational competencies. Another distinctive feature of the model is the integration of the Return on Expectation (ROE) approach, to measure not only the effectiveness of training but also the ability to align training investments with the organization's strategic goals. The model also introduces a system that promotes a continuous learning cycle, in line with the Lisbon Strategy’s goals of supporting lifelong learning. Finally, the document presents an extension of the model through the integration of predictive analytics as an anticipatory process to identify future needs and facilitate alignment between existing competencies, those required within the organization, and those demanded by the market.