Aditya worked on improving the stability and observability of the face-swapping feature in the hacksider/Deep-Live-Cam repository. Using Python and leveraging skills in computer vision and error handling, he addressed a runtime crash caused by missing face embeddings by introducing explicit defensive checks in the face detection workflow. He also enhanced error reporting by replacing ad-hoc print statements with structured logging, making debugging and maintenance more efficient. Although the work focused on a single bug fix rather than new features, the changes reduced downtime and improved the reliability of real-time face recognition, reflecting a thoughtful approach to robust software engineering.