Vineet Bansal: Bringing computing expertise to machine learning, statistical research

Written by
Sharon Adarlo
April 22, 2020

At Princeton University’s Center for Statistics and Machine Learning (CSML), faculty and other researchers can tap a variety of resources to boost their projects. Whether finding sources of funding for a study or learning the newest machine learning techniques to scale up their experiments.

Some of them have turned to Vineet Bansal as a valuable source for help. Bansal is a senior research software engineer who jointly holds office at CSML and the Princeton Institute for Computational Science and Engineering (PICSciE). His job is to amplify research with his software engineering skills.

Since he arrived on campus in late 2017, Bansal has collaborated with several faculty on projects that have ranged from building a web portal to display the behavior of various proteins and to upgrading and refactoring code so that programs can crunch numbers faster.

“It’s been wonderful working here,” said Bansal. “I get to work on interesting software projects that cover a wide variety of disciplines. No two projects are the same.”

Peter J. Ramadge, CSML director, said Bansal has been an essential part of the community because these projects would not be possible without his help. “The projects Vineet has worked on also embody important parts of the center’s mission: fostering a spirit of collaboration and cross-disciplinary research on campus,” said Ramadge.

Bansal is one of seven Princeton research software engineers assigned to different centers or departments on campus. Before starting to work on projects, Bansal and Ian Cosden, director of Research Software Engineering for Computational & Data Science at PICSciE, meet with different faculty members. Along with CSML’s input, Bansal and Cosden go over proposals from diverse faculty who want help on software engineering solutions. 

“As a software engineer, my area of expertise includes dealing with a lot of different domains and different programming languages. An important part of my job is how to put these pieces together so they can last a long time,” said Bansal.

He has worked on a machine learning project to analyze genome sequences with Ryan Adams, professor of computer science and director of CSML’s Undergraduate Certificate Program in Statistics and Machine Learning, and with Matthew Salganik, professor of sociology and interim director of the Center for Information Technology Policy, on Princeton’s Fragile Families & Child Wellbeing Study. In addition to other faculty, he has collaborated with Brandon Stewart, an assistant professor of sociology and the Arthur H. Scribner Bicentennial Preceptor, and Peter Melchior, an assistant professor, jointly appointed to CSML and the Department of Astrophysical Sciences.

Before coming to Princeton, Bansal was a software development engineer at Brooks Instrument in Hatfield, Pennsylvania. At the company, Bansal programmed and tweaked physics-based computer models, and developed applications to maintain and visualize data collected during the company’s research projects. Bansal also served as a senior software developer at Bank of America, where he was part of a team that deployed Python on a liquidity risk management project. He was also a software developer at Michigan State University’s Center for Language Education and Research, where he worked on creating interactive language-learning software for the center and developed and maintained distance-learning web applications.

One recent Princeton project under his belt involved a collaboration with Amit Singer, professor of mathematics and the Program in Applied and Computational Mathematics. This work was towards the development of a Python version of a software package called ASPIRE (Algorithms for Single Particle Reconstruction), which analyzes data from electron microscope images of molecules held in quickly-frozen ice. To read more on that project, see this article.

Bansal has also worked with Mona Singh, professor of computer science and the Lewis Sigler Institute for Integrative Genomics. Her lab specializes in computational molecular biology and heavily uses machine learning and algorithms as part of her research. 

With Bansal’s help, the lab released a web portal (found here) that can analyze protein sequences and display results. For this project, Bansal worked on creating a unified framework that combined several projects, said Anat Fuchs, a graduate student who is part of Singh’s lab at the Lewis-Sigler Institute for Integrative Genomics. In his work, he was responsible for learning about four different projects in Singh’s lab, getting familiar with their code-base, and consolidating them into a single web-based application. 

“It was beneficial for us as researchers to learn from Vineet new engineering practices that can improve our code and the way we design our computational tools,” said Fuchs. “I also really enjoyed teaching Vineet about our scientific domain of protein interactions and the motivation behind our research. I’m sure Vineet’s work will increase the visibility of our research projects and enable greater public access to algorithms and tools that are valuable to the scientific community.” 

When the website was published, Bansal said he felt a sense of satisfaction.

“It was wonderful to see it in action,” said Bansal. “It was a true team effort, and I was happy to be part of it.”