BITS Pilani Students Build AI “Research Copilot” To Help Scientists
Share
Two students at BITS Pilani’s Hyderabad campus have built an AI tool meant to help scientists with one of the hardest parts of research: working out what to try next.
Arnav Kulshreshtha and Sattwik Das started NovAtom Labs, a student-led deeptech startup, while working on freelance AI projects and their own research out of their hostel rooms. Their tool, which they call a “research copilot,” is designed to sit alongside researchers and help them understand a problem, go through the existing scientific literature, keep track of experimental limits and past results, and suggest promising directions for further study.
According to the startup, researchers linked to ETH Zurich, Georgia Tech and the University of Toronto have used and tested the technology. In India, it has been demonstrated to researchers and engineers at the Semiconductor Laboratory (SCL) in Mohali, and NovAtom Labs says it is now working with them to apply it to semiconductor research problems. Researchers at BITS Pilani Hyderabad working in semiconductors, materials, biology and engineering have also used it.
The idea is to support the full cycle of research, not just answer questions. Scientists often have to read through large amounts of published work while also juggling lab constraints, simulation results and physical parameters. The founders want their system to help at every stage: understanding the problem, reviewing what is already known, picking a direction, running experiments or simulations, learning from the results, and deciding what to do next.
They believe this could be useful in any field where researchers have to test a large number of possibilities, whether different materials, processes, conditions or hypotheses. That includes semiconductors, chemistry, biotechnology, pharmaceuticals, energy and advanced manufacturing.
Kulshreshtha and Das say their early work involved simply reaching out to researchers and institutions to understand how AI could help with real scientific problems.
Their long-term goal is ambitious. “We aim to develop self-driving research laboratories in India,” the co-founders said. They described these as places “where AI systems can work alongside scientists to analyse knowledge, design and prioritise experiments, learn from experimental outcomes, and continuously help determine the next direction for research.”
For the two students, it’s also about building advanced research technology in India rather than depending on tools made elsewhere.
