The unprecedented methodology has proven more efficient than traditional models and received approval from the scientific community
Also presented at the International Conference on AI in Finance 2023, this Tuesday (28) in New York, the study "Enhancing credit risk reporting using LLMs: A integration of Bayesian Networks and supervised prompting" was conducted by the Traive team and demonstrated a remarkable achievement: the reports generated by Traive's AI were considered, in 90.2% of cases, as good as or superior to those produced by credit analysts. Both materials, created by artificial intelligence and written by credit analysts, were evaluated by a group of decision-makers who did not know who had produced each content.
The research was examined by the prestigious Association for Computing Machinery (ACM). The results obtained and the endorsement received from the scientific community – the papers will be disclosed at an event that brings together scholars from the world's top universities, such as Oxford and MIT (Massachusetts Institute of Technology) – show that Traive is at the forefront of AI use. Applied to agricultural credit analysis, Traive's technology has created a unique and relevant product for the global agricultural chain.
The scientific papers were produced under the leadership of Aline Oliveira, in conjunction with Traive's Vice President of Data Science, Research, and AI Development, Mohammad Ghassemi. Additionally, the publications had the collaboration of Ana Clara Teixeira, Vaishali Mahar, and Hamed Yazdanpanah, technology professionals. The paper is available in full for reading here.
