– Formulated and developed a custom large language model chat-bot to answer client specific questions using LLAMA, Snowflake, and retrieval augmented generation, saving division 700,000 dollars.– Engineered a cutting-edge operating system for transitioning government paperwork to digital forms, utilizing Azure, Snowflake, and OCR for PDF metadata extraction, with automation reducing form generation time by up to 50%.– Led team of 4 engineers to create LLM based solution for automating client transaction generation from invoices

– Worked in the Financial Crimes Unit– Conducted comprehensive data quality assessments on fraud detection models using SQL and Python, analyzing over 110,000 unique data points to ensure data integrity and robustness.– Directed the automation of financial crime and anti-money laundering machine learning model validation using Python scripting, reducing task completion time by 90%.– Spearheaded optimization of model risk management, enhancing model governance and validation processes for clients.– Facilitated client meetings and presentations, translating complex technical findings to actionable insights.

– Conceived and executed the development of a social fitness app, Spot, utilizing Swift, enhancing user experience with an immersive UI and a robust Firebase backend.– Got selected to partake in UVA Darden’s incubator program and raised $16,000 in funding.– Successfully launched the app, achieving a user base peak of 3,000 and establishing a 15-person ambassador program spanning nationwide.

-Created datasets and trained models to improve summary accuracy by ~30%-Wrote script to assign interesting values to videos in order to increase customer satisfaction and understand vision API more in depth-Created data scraper to compare analytics on Skimli generated videos vs human summary videos across a host of different websites-Created a visual UI page for website display using React Native-Reorganized database

- Planned and presented educational town hall meetings on AI with over 1000 attendees- Worked to advocate for policy reform on the usage of facial recognition technology to assess prisoner risk in the judicial system- Worked closely with local, state, and federal officials to push for policy changes- Contributed to the passage of local ordinances restricting surveillance in Minneapolis and New Orleans - Helped facilitate advocacy efforts for more than 10 pieces of federal legislation- Helped shape the White House’s Bill of Rights for an Automated Society and President Biden’s executive order on AI

- Conducted research on integrating machine learning techniques with school interfaces in order to enhance the learning experience for students and modernize teaching.- Created an ML model using Scikit Learn Python, Numpy, and Pandas to analyze and predict student success based on academic and personal factors from a dataset of over 30,000 studentsPublished Paper: https://journals.gmu.edu/index.php/jssr/article/view/3165