Alfred University graduate student Naimul Haque has created a start-up company that utilizes artificial intelligence to ease the process of applying to graduate school. He hopes the experience he gained in the AI space while collaborating with undergraduate mechanical engineering student Jesse Buck will help him grow his fledgling venture.
Naimul is pursuing a master’s degree in electrical engineering from Alfred University and is expected to earn his degree this summer. Last September, at the suggestion of Xingwu Wang, professor of electrical engineering, he took a short course on machine vision which, incidentally, Jesse was also taking. The course, provided by EPRI—an independent, non-profit research and development organization—was offered as part of a workforce development initiative supported by the New York State Energy Research and Development Authority (NYSERDA). It taught the use of machine vision (a type of AI) to solve problems in electrical power systems.
Naimul said Jesse, a junior from Canandaigua, NY, told him about a company-wide competition, “AI Innovation Challenge,” being held at Advizex, where Jesse’s father, Stephen Buck, is a sales account executive at the firm’s Rochester location. Advizex provides IT infrastructure and enterprise application solutions, specializing in adaptive infrastructure, intelligent operations, artificial intelligence, cloud-managed services, and strategic optimization.
“The competition was for teams to find ways to use AI to analyze the sales leads,” Naimul explained. “The company gets lots of leads. You need to process them and decide which ones to pursue.”
Naimul—who has a bachelor’s degree in computer sciences and engineering from Ahsanullah University of Science and Engineering in his native Bangladesh, and also a master’s degree in computer science—is a teaching assistant in the Battery Machine Learning Lab in Alfred’s Inamori School of Engineering. The lab, funded by a National Science Foundation grant secured by Kun Wang, assistant professor of materials science and engineering, supports a training program focused on battery degradation and useful life prediction. The program, for which Kun Wang serves as principal investigator, will help prepare engineering students with skills and knowledge needed to succeed in the battery and energy storage job market.
As a classmate in the EPRI class on machine vision, Jesse knew about Naimul’s experience and interest in artificial intelligence and approached him about writing an AI program for the Advizex competition. Jesse provided him with the data set of the Advizex sales leads—which included information such as company names, what the company does, the size of the company, and who at the company provided the lead—and Naimul created an AI program to analyze them.
To complete the sales lead analysis, Naimul used multiple large language models (LLM) agents, which are AI systems (like Chat GPT) trained on vast amounts of text data to understand, generate, and predict human language. In particular, he utilized open-source LLMs Llama and LangChain to conduct his analysis of the sales leads.
“It took into account who the lead came from, and if that person has any decision-making authority,” Naimul said, explaining that the program reached conclusions on what sales leads should be pursued. He said he worked on the program for about a week over the winter. The Advizex Sales Bot team that used his program for the company-wide competition wound up finishing tied for first place. Advizex said the AI solution Naimul designed has helped sales teams focus their efforts and is projected to drive a 25-percent increase in ROI compared to traditional methods.
“The reason I was able to complete (the program) so fast was I used the same technology that is in my thesis, so it was known to me,” Naimul said. His thesis, “AI Co-pilot for GE Vernova ADMS,” studies how artificial intelligence can be used in the Advanced Distribution Management Solutions (ADMS) system in the GE Vernova Advanced Power Grid Lab. The lab, set up in McMahon Engineering Building, is part of the NYSERDA- and GE Vernova-supported workforce development initiative aimed at preparing students for careers in the growing renewable energy industry.
Naimul said he was somewhat surprised his program finished at the top in Advizex’s company-wide competition. “I’m a student. The other teams were made up of a lot of professionals with more experience.”
He credited Xingwu Wang’s mentorship in leading him to take part in and succeed in writing the AI program for Advizex. “He deserves a lot of credit. He supported me and Jesse and encouraged us a lot.”
Naimul said he expects to work in a temporary, part-time remote internship with Advizex—helping create a sales bot using AI technology—while he completes his master’s degree studies at Alfred this summer. After graduation, he hopes to raise funding for his start-up company, Gradmate, which offers an AI platform that prospective graduate students can use while choosing and pursuing a graduate school.
“The platform would help them get to know the different grad programs based on specific research fields and academic programs,” he explained. “They could use AI to put together their application materials. It would make the application process a lot easier.”