Laura Uguccioni

Laura Uguccioni is an alumna of QMSS. For her master’s thesis, she applied NLP techniques to Twitter data to explore the relationship between public spaces in New York City and well-being. Following her studies, she embarked on a career as a data scientist, joining KPMG’s AI Analytics & Engineering consulting practice. Most recently, she served as an Associate Director at KPMG’s Data Science Center of Excellence, specializing in NLP and probabilistic models. Currently, she is on an extended maternity break, cherishing time with her baby, studying Spanish while in Mexico, and working on personal data science projects.

GRE Writing Tutoring App - Brief project description

The project, the GRE Writing Tutoring App, seeks to harness the capabilities of Language Models (LLMs) to assist prospective graduate students in preparing for the GRE Writing exam. The primary objective is to create a tool that provides graded feedback on writing tasks, tailored specifically to the criteria of the GRE Writing exam. The overarching goal is to democratize access to GRE preparation resources, particularly for individuals who may not have the means to afford expensive training programs.

This project involves building an LLM App using LangChain. Some key components include prompt engineering, building AI agent workflows, evaluation, and building a demo. Based on evaluation results, fine-tuning approaches and costs will explored.