How Generative AI is Transforming the Evaluation of Online Training Courses

Every year, thousands of Spanish teachers apply for online training courses offered by the National Institute of Educational Technologies (INTEF), the body responsible for the digital transformation of education in the country. Until recently, processing these applications was a painstaking task. Certificates had to be checked manually, handwritten documents deciphered, and scores assigned according to precise criteria. This process took several weeks and required entire teams to be mobilized.

What sets GenAI EDU apart from a simple document extraction tool is its approach. “The processing of unstructured information is based on generative AI models that enable advanced contextual analysis: semantic extraction, entity and relationship identification, and summary generation,” explained Raquel Gómez Querencia, a consultant at Inetum. The solution thus goes beyond the limitations of traditional approaches based solely on optical character recognition (OCR) or fixed static rules.

For document recognition, including handwritten documents, the platform relies on Azure Document Intelligence, a pre-trained service that requires no specific configuration. The system returns confidence metrics that can be used to validate the quality of the processing.

However, automation does not mean that the machine makes decisions on its own. Administrative rules are translated into instructions that AI applies to documents.

“AI analyzes the context of documents by applying these rules and generates a reasoned response, explicitly indicating which page of the document the information comes from and on what basis it calculates the score,” explained Raquel Gómez Querencia.

Human evaluators use these results as a decision-making aid. “AI does not replace human intervention, it amplifies the user's analytical capabilities,” she added.

Building on this success, Inetum is continuing its collaboration with the Spanish Ministry of Education on other projects aimed at simplifying administrative processes in the sector. The goal is to extend the solution to other use cases in the field of e-learning.