Use of Artificial Intelligence in Istat to improve the customer experience

Abstract

The ongoing digital transformation, the evolution of computational capabilities and the widespread use of artificial intelligence (AI) make information management and intelligence strategies increasingly central themes. Starting from the state of the art, the aim of this paper is to explore how these phenomena are interpreted in Istat, through a project dedicated to the use of AI for the data search in Istat’s data warehouse, where official statistical data are published. They include, among others, national accounts, household economic conditions, prices, labour and wages, industry, justice, services, health, education, culture, welfare, daily life, public administrations, environment and energy, and agriculture.

The obtained result was to facilitate data searches through the use of an AI-based chatbot. Our goal is to  enhance the user experience by assisting users  in searching for content already published in the data browser.

Istat, as the official producer of statistics for our country, pays particular attention to the accuracy and quality of the data released. Even if the use of AI is widespread, from a business  prespective, it is not yet a mature technology. On the one hand, therefore, the drive for innovation has led us to experiment with this new technology, and Istat is among the first public administrations to offer a similar service. On the other hand, the project, despite being online and validated by the methodological sector, and despite having trained the data only on the Istat domain, has an experimental nature. What we obtained is the first step. We expect to continuously improve our data research by investing in AI.

Finally, we want to emphasize the teamwork that underpinned this project. The IT, communication and methodological departments have worked in a truly synergic way, seeking together a point of balance in the complexity, in the attempt to provide a better public service.