Enhancing Observability with Generative AI and Large Language Models: Centralizing APIs and Documentation for Improved Support Team Responsiveness; Design and Impact: Evaluating SBA Programs Efficacy in Small Business Success

Sareini, Ali, School of Engineering and Applied Science, University of Virginia
Forelle, MC, University of Virginia
Vrugtman, Rosanne, EN-Comp Science Dept, University of Virginia
Morrison, Briana, University of Virginia

The complex IT infrastructure of Fidelity Investments was characterized by thousands of APIs and a large volume of documentation and logs. To tackle the challenges of managing and enhancing observability, we proposed and developed a solution leveraging Large Language Models (LLMs) and Generative AI (GenAI) technologies. The core of our approach involved design and implementation of an intelligent system that centralizes and processes the myriad of APIs, documentation, and logs, subsequently presenting them in a coherent, accessible manner to the support team. We utilized advanced AI and machine learning techniques, requiring a comprehensive set of programming skills, tools, and understanding of AI-based text processing and data integration methods. The preliminary results of our implementation demonstrated significant improvements in the efficiency and effectiveness of the support team's operations by providing enhanced search capabilities and actionable insights into system performance issues. Further work is needed to refine the system's accuracy in document and log interpretation, expand its API coverage, and improve its user interface. Future efforts will also focus on extensive testing and evaluation to identify and rectify any bugs or glitches, as well as explore broader applications of this solution across different departments within Fidelity Investments to maximize its organizational value.

BS (Bachelor of Science)
AI, GenAI, NLP, SBA, Government, Site Reliability, Software Engineering

School of Engineering and Applied Science
Bachelor of Science in Computer Science

Technical Advisor: Brianna Morrison
STS Advisor: MC Forelle

Technical Team Members: Ali Houssain Sareini

Issued Date: