Design and Evaluation of Techniques for Enhancing the Utilization of Pre-Trained Language Models (GPT-3) in Sales and Marketing through Prompt Engineering and Fine-Tuning

This research project is focused on using AI technology called Pre-trained Language Models (PLMs) to generate results that are more human-like and of higher quality. The team will utilize a technique called “prompt engineering” to work with a leading PLM in the market, GPT-3, to enhance productivity in reading and writing tasks for salespeople. By collecting data from business users, the team will fine-tune GPT-3 to perform specific tasks that salespeople currently handle. The ultimate goal is to enable the company, Robust Choice, to incorporate GPT-3 into its products and provide better value to its clients, ultimately saving them time and money. Additionally, the project aims to innovate, develop, and commercialize a prompt pipeline that utilizes various methods to create unique prompts for specific sales tasks.

Faculty Supervisor:

Mohammad Hassanzadeh

Student:

Partner:

Robust Choice

Discipline:

Computer science

Sector:

Information and cultural industries; Professional, scientific and technical services

University:

University of Windsor

Program: