Research Analyst, LenzVu
Apr, 2024 - Present
Engineered a new AI feature for generating responsive HTML emails, developed RESTful web services, and executed ETL processes.
Engineered a new AI feature that generates fully responsive HTML emails with a single click, utilizing technologies such as Node.js, Express, TypeScript, and typeORM (SQLite).
Conducted extensive research on utilizing Large Language Models (LLMs) for generating HTML emails, enhancing expertise in advanced prompt engineering and LLM evaluation.
Developed a RESTful web service using Python and FastAPI, to evaluate and benchmark multiple LLMs including OpenAI GPT, Gemini, and various open-source models from hugging face, improving model selection process.
Developed web scraping service using OpenCV and Selenium to extract data and analyze color/shape attributes of call-to-action buttons on websites, resulting in a 20% improvement in the quality of training data.
Designed and executed ETL processes with Azure Data Factory and Azure SQL Database to extract, transform, and load email data, facilitating the fine-tuning of LLM, improving model accuracy by 45%.