Deployment Strategist

Selected as one of the Top 10 MBAn Capstone projectsEnhanced RAG-based chatbots by improving document retrieval accuracy and the quality of text generation, while automating test set creation to reduce the process from 8 hours to 5 minutes. Developed an automated system for evaluating LLM-generated marketing texts, cutting manual review time by 60%.

Co-organized world’s largest fully student-driven AI/ML conference, with 400+ attendees

Research Assistant to Professor Jonas JonassonApplied XGBoost to stratify patients by opioid overdose risk, leading to targeted interventions that reduced overdoses by 50%. Conducted an impact evaluation using causal inference to assess the effectiveness of the intervention program, driving data-backed improvements in healthcare outcomes.

Enhanced ClassHook's video categorization with advanced NLP and large language models, leveraging Hugging Face transformers for proficient tagging of 7,000+ videos into 100+ subject categories; see Projects section for more information

Drove the Disaster Recovery initiative for AI Infra training jobs and accomplished a reduction of compute wastage time by 30% through scheduling optimizations and analysis of current training model architecture

Conducted research for Professor Michael Littman on a funded DARPA project for the “Lifelong Learning Machines” (L2M) program

Led the Sustainable AI initiative by estimating the cost and carbon footprint of compute, storage and networking infrastructure for training and inference of ranking models

Developed an educational web platform for Kirkland Financial Group, featuring financial webinars, seven comprehensive courses, and over 25 lessons, tailored to support user engagement and client budget management

Collaborated on an AI-driven mentorship platform, leading research and development efforts to leverage machine learning in matching over 10,000 mentors and mentees, and contributing to the creation of a search ranking algorithm for optimal pairings

Engineered a dual-functionality web platform 'junk.ai', leading the design of a text classification application and implementing an image recognition system with machine learning to enhance digital content curation