CRIF’s hub and practical use cases in the financial sector: document management automation and control automation
The rapid acceleration in the use of Generative AI is radically transforming the financial sector. Banking and insurance players, aware of the huge potential of this technology, are implementing AI-based solutions to increase operational efficiency and deliver cutting-edge services to customers. As one of the first to adopt GenAI in the data industry, CRIF stands out due to its industrial approach to the integration of artificial intelligence.
A cross-country approach to GenAI
CRIF has centralized its GenAI initiatives within its InnovEcoS innovation hub, whose mission is to drive the development of business models and future technology trends across borders, innovating services in full compliance with legal and organizational aspects. “This approach has allowed us to develop solutions that integrate GenAI directly into existing processes and software in order to centrally manage the creation of prompts tailored to individual use cases, associating them with the calculation of KPIs,” explained Pietro Curtolillo, CRIF Marketing & Communication Director.
Two real-world cases of innovation
A significant example of CRIF’s innovation is its document management automation, which combines retrieval augmented generation and prompting techniques to automatically explore and annotate a broad document base and incorporate the results into business processes. This tool has transformed the manual extraction of KPIs from text documents (financial statements, real estate valuation reports, etc.) into an efficient and accurate automated process, allowing operators to easily compare and validate the automatic transcription results. CRIF followed the same principles when developing another GenAI-based tool that supports second-level controls in banks. Using large language models and specific prompts, CRIF has automated controls on loan dossiers, allowing 100% of cases to be analyzed instead of just a sample. This system performs an automated pre-assessment that identifies the reasons behind detected anomalies, greatly improving the efficiency of the control process and reducing the time taken. As demonstrated in real-life cases, this methodology can reverse the 80-20 rule on the time taken to search for information versus the analysis time, allowing operators to focus on anomalies.
Data quality and skills to seize opportunities
“The adoption of Generative AI in the financial sector offers huge opportunities but also presents significant challenges, such as the need for high-quality data and specialist skills for these initiatives to succeed. CRIF is aware of the potential ethical risks associated with the use of AI and is committed to developing robust policies in this regard. In the short term, the adoption of GenAI by financial players will focus on the automation and efficiency of existing processes; a transitional phase toward a future in which products and services will be natively based on generative technologies,” concluded Curtolillo. With its practical and value-creation approach, CRIF is ready to guide its clients toward this new era, where the watchwords are efficiency, security and innovation.
