This senior thesis project focuses on developing an AI-powered static analysis tool designed to detect security vulnerabilities in C and C++ codebases. Leveraging machine learning and advanced pattern recognition techniques, the analyzer scans source code to identify common and subtle programming flaws, including buffer overflows, use-after-free errors, and unsafe pointer operations. By combining traditional compiler-driven analysis with neural network-based learning models, the tool aims to assist developers in writing more secure, robust, and maintainable code. The system will be evaluated against known vulnerable code samples and benchmarked for precision and recall.
The PoC or proof-of-concept folder stores the prototype of the main project. In this folder the user can find a prototype version of the Ai-powered static analysis tool that I have developed to showcase and prove the concept of my senior thesis for my professors and readers.