- Delivered key updates, including conversational prompting for the AI chatbot using TypeScript- Utilized the OpenAI Node API Library and TypeScript to engineer one-shot prompts, categorizing AssemblyAI transcripts into “sport,” “student,” “key takeaways,” and “drills” with GPT-4 for semantic analysis- Improved backend data consistency using a serverless cloud function in GCP/Firebase- Managed app builds for App Store release and for QA testing via TestFlight, including permissions,configuration, bug tracking, and submitting fixes- Owned a complete UI overhaul, implementing theme functionality to support UX and brand experimentation with TypeScript and React Native with updated colors, fonts, and layout

- Developed coursework, taught selected classes, and supported Professor Lori Levin by grading homework

- Trained Siri NLP models to enhance contextual awareness, refine phonemic translation rules, and strengthen linguistic modeling- Owned the design and development of the Siri Danish phoneset, syllabifier, and phonemic translator from English to Danish to support Text to Speech and Speech to Text, developing translation mappings and sequential processing rules for sounds in Python- Reduced QA timeframes and runtime by 3X by optimizing a long-running Python script (2+ hours) to prevent hangs. Automated request resends for timeouts and errors, implementing a 5-second wait rule to enhance reliability- Developed a Python automation script to compare Phonemic Translator rule outputs against gold standards, replacing manual assessments by linguists and yielding insights into rule structure and quality

- Trained Siri NLP models to enhance contextual awareness, refine phonemic translation rules, and strengthen linguistic modeling- Owned the design and development of the Siri Danish phoneset, syllabifier, and phonemic translator from English to Danish to support Text to Speech and Speech to Text, developing translation mappings and sequential processing rules for sounds in Python- Reduced QA timeframes and runtime by 3X by optimizing a long-running Python script (2+ hours) to prevent hangs. Automated request resends for timeouts and errors, implementing a 5-second wait rule to enhance reliability- Developed a Python automation script to compare Phonemic Translator rule outputs against gold standards, replacing manual assessments by linguists and yielding insights into rule structure and quality

- Developed course content for the technical bootcamp for Intro to NLP, Language Models and Smoothing, Classification, and Deep Learning Modules- Designed Python programming projects in Jupyter Notebook to support student learning, including working with PyTorch and TensorFlow, developing Naïve Bayes Classifiers, using spaCy for text classification, one-hot vector representation, word embeddings, and classification with scikit-learn

Graded programs for C++ Advanced Programming class & Compiler Design class

Designed and implemented English analysis using Python