The project for Masaf aims to provide an automated system for document analysis and summarization. This system uses large language models (LLMs) to extract key information, generate summaries and create analysis reports.

Main Features:

• Implementation of an agent for reading documents and creating reports
• Automatic checklist completion
• Tools for analysis and verification by users

Document analysis

The system analyzes the provided documents by single scanning them and progressively integrating the extracted information.

At the same time, a checklist that contains a predefined list of relevant information is processed. Meanwhile, summarization algorithms generate a discursive summary highlighting the salient points that emerged from the analysis of the documentary material.