Supported over 600 students during office hours with questions on programming and assignments. Graded around 80 submissions per week and helped design coursework focused on Java and object-oriented programming. Worked across a broad set of topics, from core programming concepts to using command-line tools.

Built a real-time speech-to-text interface using WebSockets and JavaScript, integrating AWS Transcribe for live transcription. Improved medical vocabulary recognition by 12% by comparing outputs from multiple services (OpenAI, Google Cloud, Amazon Transcribe, Assembly AI). Since its release in June, precision and recall have improved by 3%. Also worked with LLaMA 2 and transformer models during development.

Analyzed four years of Walmart’s supply chain data using SQL and a network simulator. Used pandas and scikit-learn for data cleaning and comparison, helping reduce cloud storage by over 5TB. Streamlined data processing workflows, which cut down simulation errors by 28% and improved run-time efficiency by 35%, helping shape the tool into an MVP for internal teams.

• Conducted research on software development for machine-human interfaces, including biometric technologies.

• Implemented a thumb print recognition feature, resulting in a 30% increase in user satisfaction.

Developed and implemented a new feature allowed users to unlock their phones quickly and securely using thumbprint recognition technology, resulting in a 30% increase in user satisfaction