Tharun Cota Builds a Vision of Innovation

By: Rebecca Moon Ruark

“I wanted to present something new, something that actually helps everybody,” said Tharun Cota of the inaugural IMPACT Symposium at Northeastern’s Boston campus, which joined hundreds of graduate students—along with entrepreneurs and software solution companies—to the Boston campus to showcase their experiential learning and practical experience.

Created as a project for INFO 7375: Prompt Engineering for Generative AI, with course lecturer Shirali Patel as mentor, Cota developed “VisionCheck: AI-Driven Idea Evaluation for Novelty and Innovation.”

It wasn’t his first idea. During his project’s discovery stage, he found that one of the early ideas he came up with already existed. So, he scrapped it. But it made him wonder, what if there was an application that could tell him if an idea was new—or not? “There wasn’t an application,” he said. “So I thought, let’s build one.”

A real world application

Cota knew a master’s degree in information systems was in his future, after earning his bachelor’s degree in computer science and engineering from Panimalar Engineering College, Anna University in India. He found what he was looking for in Northeastern, Arlington’s Master of Science in Information Systems program, an education that was both broad and deep; a holistic approach, from front end to back end, from data science to AI, and from software development to business analysis. “If you can’t sell it, what’s the point of building it, right?” he asked.

In his second year at Arlington, Cota set his sights set on presenting at Boston’s IMPACT Symposium, which provides students a platform to present their practical and experiential work and demonstrate how their efforts contribute to industry advancement and societal impact. Open to all graduate students at Northeastern University, the symposium also invites faculty members, industry professionals, and potential employers who are interested in learning about the innovative projects and research being driven by today’s students.

Cota’s project, VisionCheck, introduces an AI-driven tool designed to evaluate the novelty of user-submitted ideas and provide actionable insights for refinement or innovation by leveraging data from patents, academic repositories, and real-time web trends. By converting ideas into vector representations and comparing them against data from patents, academic repositories, and real-time web trends, the system generates a novelty score and delivers personalized feedback to help users refine, innovate, or pivot effectively. “Every lecture, Dr. Patel would teach a concept, and I would try to implement it on this application,” Cota said. “It helped me to learn the subject and to use it, practically.”

The real test for VisionCheck came at the symposium, where presenters have the chance to build their professional networks and enhance their visibility with potential employers. More than a poster project, the developed application of Cota’s provided feedback on a wide range of ideas, allowing participants to experience results in real time. Wildest idea tested on VisionCheck? “One person wanted to create an Ironman suit that flies over short distances for food delivery,” said Cota. The result? The application gave the idea a fully novel score—with feedback including a research paper detailing the building of a superhero suit and a couple related patents. “It was a cool idea,” he said. “Maybe one day we could team up!”

The AI application was well-received at the symposium, where Cota was able to exercise his networking skills. “Founders liked the idea,” he said. Feeling encouraged that his project has a lot of scope, he plans to pursue it further. “Info retrieval is trending in the field of AI,” he said. “Everybody is trying to validate solutions and retrieve information in the quickest way possible.”

Not one to be content with the status quo, he has already implemented feedback received from members of industry at the symposium, applying security aspects and a retrieval-augmented generation framework for less lag in retrieving sources. “It’s a project that’s very close to my heart,” he said. “It really shined.”

Learning to teach, teaching to learn

Nearing graduation, Cota looks back on his time at Northeastern, Arlington. He counts serving as a teaching assistant for INFO 6105: Data Science Engineering Methods and Tools as one of his greatest accomplishments over the two-year program, “helping each student to get the most from this course—and to flourish,” he said.

He wanted to assist for this course, in particular, because he had struggled with it in his first semester, and he knew the “pinch points.” Initially, he thought his role would consist mostly of grading for course instructor William Claster, Associate Program Director – and Associate Teaching Professor, but he found that being a teaching assistant was much more hands-on—specifically, learning to help a group of people of diverse ages, backgrounds, and industry experience.

His first task was to develop alignment, a neutral starting point across all the students, before building shared knowledge. Rather than meet remotely, Cota made a point of meeting with students in person to gauge their progress. “It was my role to ensure the students understood and fell in love with the course, like I did,” he said.

For his dedication and exceptional achievement, the College of Engineering awarded him the 2024 Outstanding Student Teaching, Service, and Leadership Award (MS).