Background
Huge amounts of sensitive data and complex information result in structured and unstructured data that must be read, classified, and masked.
Public Sector
Data classification and anonymization
Solution
The NLP- and NEE-based tool enables it to read files of different formats and extract, through OCR technology, specific data that are then anonymized.
Results
The process of data extraction, classification and anonymization has been speeded up through task automation, and the margin of error minimized.
Chatbot for legal users
Background
For legal professionals, researching large amounts of documents in one’s possession (e.g., judgments, court papers, etc.) is a complex job, requiring scrupulous attention to sources and therefore very long time.
Solution
The Virtual Assistant, developed according to Retrieval AugmentedGeneration methodology and based on Generative AI (Large Language Models),knows how to respond to the user on everything related to the dataset used (e.g., court papers, judgments, etc.) and can also return, for each response, the detailed list of sources used.
Results
Users have quick access to information and get articulate and complete answers, documented with a list of the sources the chatbot has drawn on.
