• Computer Science at Georgia Institute of Technology
• Interested in Machine Learning, AI Engineering, and Software Engineering
• Building systems around LLMs, ML evaluation, retrieval, and developer tooling
• Focused on turning ML systems into reliable, deployable products

AI-based Multilingualization Quality Improvement · Summer 2025

• Built a multi-stage machine translation quality assessment pipeline combining semantic similarity, COMET-based evaluation, rule-based validation, and selective LLM assessment.
• Evaluated translation quality across 300K+ multilingual strings and incorporated over 1,000 human-reviewed translation pairs into the evaluation process.
• Reduced human QA review workload by 87% while improving translation quality assessment performance by approximately 67%.
• Received Runner-up at Fasoo Fix Day for the project.

Innovation Management — Data Dashboard / Tableau Copilot

• Developed an AI Copilot for Tableau dashboards that enables natural-language Q&A over research administration data.
• Built Python pipelines to parse Tableau TWB/XML metadata and structure dashboard information for downstream analysis.
• Implemented retrieval and Q&A workflows using embeddings and RAG to connect user questions with dashboard data and metadata.
• Developed logic for automatic visualization generation and dashboard analysis from CSV/Excel and Tableau data sources.
• Worked with research administration datasets covering sponsor funding, departments, principal investigators, and funding allocations.

AI-assisted pull request risk analysis and triage system built around a fine-tuned code model, deterministic static analysis, and controlled evaluation.

Georgia Tech student assistant using controlled agentic RAG, hybrid retrieval, structured course data, and citation-grounded responses.

Developer-focused tooling for improving programming and software engineering workflows.