The National Institute for Insurance against Workplace Accidents (INAIL) is an Italian public administration responsible for the management of mandatory insurance against workplace injuries and occupational diseases. The main role of the institute is to protect workers by offering prevention services, healthcare assistance, compensation, and reintegration support for those who suffer injuries or contract diseases related to their work activities.

In 2024, INAIL faced a significant challenge in the implementation of its funding program, “Bandi ISI”. The new edition, published in December 2023, saw a 65% growth in the budget—from EUR 300 to 500 million. This increase led to a 50% rise in applications from companies, placing considerable pressure on the management system.

To ensure the sustainable management of the new initiative, the foundation decided to support their contact center and offices with an artificial intelligence (AI) virtual assistant capable of helping companies with the submission and admission processes. This support would be available directly through the web application used to access the procedure, reducing the need for ticket submissions or calls to operational support structures.

INAIL collaborated with IBM Consulting® and IBM® Client Engineering to assist its contact center and offices in overcoming possible delays and managing the anticipated surge in requests. The team launched a pilot project and created a virtual assistant capable of answering questions about the funding application process.

With the help of the assistant, they can now seek information about the call, eligibility requirements, and operational procedures. Additionally, they can standardize responses provided by INAIL’s contact center, where staff handle more specific inquiries.

The pilot ran for 2 months with people from both INAIL and IBM, using unstructured data such as manuals, FAQs, previous tickets or requests and official documentation. The solution involved IBM watsonx Assistant™ conversational AI solution, which was able to understand and address various questions, general or specific, by leveraging multiple features and acting as the orchestrator between IBM Watson® Discovery platform and IBM watsonx.ai™ studio.

IBM Watson Discovery was used as a documentation collector and cognitive engine available to the overall application. As a result of Watson Discovery’s indexing functionality, the system was able to identify portions of text where the answers could be found and then manage them through watsonx Assistant and watsonx.ai. Finally, the gen AI Mistral LLM model (mixtral8-7B) on watsonx.ai was used to engage the conversation and generate answer texts based on manuals and documentation that contained the correct and relevant content.

Overall, INAIL hopes that the virtual assistant can provide companies with timely guidance during the funding application phase, easing the pressure on the contact center and allowing staff to focus on more complex issues.