Growth, Treasury, Strategy

Advanced Analytics & Machine Learning - R&D | Manufacturing | Consumer Brands

• Participated in the design and development of learning analytics applications to enhance teaching and learning in theclassroom & remote.• Developed speaker transcription & diarization data pipelines using GCP's SpeechClient API to supplement thefunctionality of Spectrum -a teaching insights application. Trained a simple XGBoost model to detect questions anddeclarative language in the lecture recordings.• Tracked and researched emerging open-source technologies in the ML and AI space and created PoC's forapplications aimed at enhancing learning experience in the university. Worked primarily with Python, R, GCP/AWS, Bash, BERT, XGBoost, GPT-2

• Utilized GPT-2 and BERT language embeddings to enhance the modeling of student questions, ensuring relevant answers were mapped to effectively guide their educational track that was implemented into a chatbot.• Designed a logic-based algorithm to map out college degree paths, offering real-time responses to student inquiries across more than 1000-degree programs using NLP architecture to understand different questions.• Accepted into Google's Research Program to advance the intersection of education and technology, leveraging funding credits to operate AIODA on the Google Cloud Platform (GCP).

Reported to the Mars Advanced Analytics & Machine Learning desk this past summer. Learning the ins and outs of CPG and Pet Care industries, my main assignment was to assist a new product launch in the UK and use my data science knowledge to define a problem and propose a viable solution.• Extracted conversation topics, sentiment, and engagement via social media platforms; discovered further range and product extensions.• Deployed data-extraction pipeline for ‘data-poor’ environments by utilizing Speech2Text solutions, now used for multiple Mars brands. • Conducted bi-weekly meetings with UK associates to provide brand specific progress reports on social media landscape and engagement.

LMS is an educational technology startup company based in State College aiming to provide first generation college students the proper resources to land their first job upon graduation. LMS is an important foundation in my understanding of NLP and using data science as a social force for good.• Web-scraped higher-educational and government repositories. Wrote regular expressions to parse and extract data for NLP models • Deployed WordNet model to extract similarities between different course and job descriptions to serve as recommendation algorithm. • Developed data quality benchmarks through MTurks; our approach was unsupervised, leaving us in the dark on recommendation quality.

• Learned the fundamentals of futures and options trading through a 10-week rotational program through different commodities desks.• Constructed algorithm to automatically align commodity spot prices for clients depending on the chat platform they were sent through.