I Agents for Occupational Safety: Training for Injury Prevention, Internal Communication and Marketing

his position involves the design, development and experimental testing of conversational AI agents — based on open-source European Large Language Models and advanced Retrieval-Augmented Generation (RAG) architectures — as the central enabling technology for occupational health and safety management. The researcher will help build layered regulatory datasets (Legislative Decree 81/2008, EU directives, sector protocols) and design context-aware retrieval strategies to develop AI agents able to: deliver personalised injury-prevention training based on the user's professional profile and competence level; support internal corporate communication on safety topics; generate multimodal (textual, visual, multimedia) marketing and awareness content. Experimental validation will involve workers from different sectors, measuring training effectiveness, communicative engagement and impact on injury-risk reduction.