Page 108 - tech fest 2025 ב״ש חוברת תקצירים

108
Evaluating the Capabilities of Large Language Models in Performing Financial Calculations
IEM-C-03
Shira Dahan; shirada7@ac.sce.ac.il Yuval Dvir; yuvaldv@ac.sce.ac.il
Advisor: Dr. Dima Alberg
SCE - Shamoon College of Engineering, Be’er-Sheva
This research project examines the capability of ‘large language models’ (LLMs) to perform loan repayment calculations accurately. Rather than developing a new model, this study evaluates existing LLMs-ChatGPT and Gemini-using prompt engineering techniques and a set of predefined evaluation criteria. The objective is to assess the models’ abilities to apply financial formulas and generate correct amortization schedules. The uniqueness of our project lies in its focus on comparing general-purpose, conversational AI tools for financial computation tasks to determine their potential usefulness for individuals with limited financial literacy. These findings should contribute to the future development of AI-based financial support systems for non-expert users.
Keywords: AI, amortization schedules, LLMs, loan repayment, prompt engineering
Analysis and Enhancement of a Hospital Temperature Control System
IEM-C-04
Liron Kobrinsky; liron.kobrinsky@gmail.com Miri Shalem; miri7950@gmail.com
Advisor: Mr. Meir Efrat
SCE - Shamoon College of Engineering, Be’er-Sheva
Maintaining temperature control in hospitals is crucial to the prevention of damage to medical equipment and guarantees optimal conditions. The existing system monitors data and detects anomalies but lacks an automatic fault response. This project enhances that system by integrating an automatic correction mechanism and analyzing data that identifies patterns and improves performance. The main challenges are processing and analyzing data to determine the causes of temperature anomalies and optimizing the control mechanisms. Our project focuses on designing an automation solution for immediate fault correction - thus, reducing manual intervention and improving system stability. Conducted in collaboration with “Afcon,” a leader in control and automation, this project improves Afcon’s control system, which currently monitors and regulates the temperature in hospitals.
Keywords: anomaly detection, automation, data analysis, temperature control