Experience
Experience
Machine Learning Engineer Intern — Adobe
Built an LLM-as-Judge evaluation framework benchmarking frontier models (GPT-4o, Gemini, Claude) for multimodal chart quality assessment. Developed a LangGraph agent with MCP tool integration for automated evaluation feedback loops, cutting manual review time significantly.
Research Assistant — Indiana University Bloomington
Applied ML research in the Luddy School of Informatics, Computing, and Engineering. Focus on LLM inference optimization and machine learning systems, complementing the MS Data Science program.
Data Scientist II — S&P Global
Built production NLP and document-AI pipelines for financial data. Delivered a sparse-retrieval RAG system with a fine-tuned T5 encoder deployed on NVIDIA Triton (HTTP) on AWS EC2. Managed experiment lifecycle with MLflow and DVC. Automated document-processing workflows at enterprise scale.
Data Scientist I — American Express
Improved call quality assessment accuracy by 10% with a VIBE model combining XGBoost with BERT embeddings. Built and deployed LSTM-based demand prediction models on AWS EC2 serving millions of cardholders daily.