Bias Busters: How Two IIT Kharagpur Minds Are Cleaning Up AI
Artificial intelligence is everywhere—powering decisions, shaping narratives, and, too often, amplifying biases. Sumit Verma and Pritam Prasun, IIT Kharagpur alumni, step up with Responsible AI Labs (RAIL) to tame AI’s wild side.
Armed with top-tier education and years of tech experience, they’re on a mission to make AI fairer, more transparent, and worthy of trust. Verma’s story starts at IIT Kharagpur, India’s renowned engineering hub, where he first grappled with AI’s potential. Post-graduation, he spent five years at leading tech firms in
Japan, a hotbed of AI innovation. But amid the breakthroughs, he spotted flaws.
“I saw how biased and unreliable AI was becoming,” Verma said. “It wasn’t just a glitch—it was a societal threat. ” That realization pulled him back to India, determined to rewrite AI’s playbook.
He found an ally in Pritam Prasun, a fellow IIT Kharagpur alum and tech industry veteran. Bonded by their alma mater’s problem-solving spirit, they co-founded RAIL to plug a glaring gap: accountability in AI. Their first weapon? The RAIL Score, a sharp new metric unveiled in preview form. It sizes up AI content for fairness, privacy, and transparency, giving developers a roadmap to fix biases fast.
But they’re not stopping there. RAIL is also crafting a Mistral-7B-based language model that ditches quick answers for something deeper—a mentor-like guide to spark critical thinking.
“We want AI to teach, not just tell,” Verma said, a nod to the intellectual rigor he honed at IIT Kharagpur. While Verma spearheads efforts in India, Prasun is taking their vision global. This week, he’s in Atlanta, presenting RAIL at DrupalCon, a major tech gathering. He’s spotlighting the RAIL Score and their broader ethos, proving their IIT-rooted skills travel far.
The timing couldn’t be better. AI’s flaws—think gender stereotypes in chatbots or racial skews in algorithms—are under fire. Verma and Prasun don’t just point fingers; they build fixes. From bias audits to ethical automation, RAIL’s solutions reflect the precision they mastered at IIT Kharagpur. For Verma, it’s personal.
“AI can lift the world up,” he said, “but only if it’s responsible.” With Responsible AI Labs https://responsibleailabs.ai/, these IIT grads are betting big on that vision. As their work gains steam, they’re showing that cleaning up AI might just start with two sharp minds from Kharagpur.
