Chemical Engineering Undergraduates Win $100,000 in AI Challenge That Rewards Real-World Progress
Teams from UC Berkeley, Baylor College of Medicine and Columbia, and Texas A&M share $175,000 in Mays Business School’s inaugural competition.
Dualizer (formerly EntroPINN), a team of chemical engineering undergraduates from the University of California, Berkeley, celebrates winning the $100,000 grand prize at Mays Business School’s AI Venture Velocity Challenge. The competition brought together student entrepreneurs from across the country to test and advance AI-enabled ventures through real-world experimentation.
A team of chemical engineering undergraduates from the University of California, Berkeley won the $100,000 grand prize in Mays Business School at Texas A&M University’s AI Venture Velocity Challenge after spending five months testing their industrial modeling venture with practicing engineers, discovering limitations in their own technology, and using AI to rebuild and test their approach faster.
Their venture, Dualizer (formerly EntroPINN), rose from a national field of 531 teams across 160 institutions. From May through September, teams advanced through successive rounds of structured experimentation, with 248 submitting experimentation logs, 27 reaching the semifinals, and 12 advancing to the finals. The challenge was designed to reward evidence over presentation polish: whether teams made contact with real customers, tested the right assumptions, and used what they learned to make better decisions. The finalists competed in judge-led conversations about their evidence, decisions, and next steps, followed by time to reflect and respond. Teams did not present slide decks.
The other winners stood out for similarly impressive progress following disciplined experimentation. NEXUS, a team of genetics doctoral students from Baylor College of Medicine and Columbia University, earned second place and $50,000 for a clinical-trial matching venture that secured a paid pilot commitment. Lobe, founded by two Texas A&M University undergraduates who combined business and technical backgrounds, received third place and $25,000 for senior-living software that already has paying customers.
Across the winning teams, the common thread was real-world contact, sharper human judgment, and faster execution with AI. Other finalists included four additional teams from Texas A&M University, along with teams from High Point University, Tulane University, the University of Silicon Valley, the University of Texas at Austin, the University of Tulsa, and the University of Wisconsin–Madison.
“We rethought what it means in an AI age for students to try to build companies,” said Levi Belnap, executive director of the Center for Applied Entrepreneurship and Innovation at Mays Business School. “We wanted students spending their time learning from customers and making progress, not perfecting a presentation. We gave teams five months to co-build real companies with AI, to talk to customers, to test the right assumptions, and to use AI to test and build faster, discover where their assumptions were wrong, and decide what to do next. The teams that rose to the top made the most progress doing that and challenging themselves. What separated the strongest was not simply how much they built. It was how well they turned evidence into better decisions.”
The judging panel at the finals included Sarah Elk, Bain & Company; Ash Maurya, LEANSpark; Mitra Miller, Houston Angel Network; Ujjwal Rajbhandari, Autonomize AI; Betty Adams, Mays Business School; Tyler Wooten, FocusAI; Jack Roehr, Houston Angel Network; Elizabeth Haegelin, early-stage investor and Texas A&M graduate; and Naisha Kumar, a senior at Centennial High School. Additional judges who evaluated teams in earlier rounds brought experience in AI and venture development from Flybridge and Harvard Business School, Google, AliveCor, Netflix, Ciroos, Meta, Calumet, and Vanderbilt University.
2026 AI Venture Velocity Challenge Winners
Grand Prize, $100,000: Dualizer (formerly EntroPINN), University of California, Berkeley
Led by Nathan Liu, Dongha (David) Kim, and Christopher Donahue
Dualizer is building AI tools to turn industrial process diagrams into engineering simulation models, helping engineers evaluate complex industrial operations.
The chemical engineering team advanced through rigorous technical benchmarking, direct conversations with practicing engineers, and a disciplined decision to narrow its scope when broader generalization did not hold. After improving its diagram-to-model conversion based on engineer feedback, Dualizer plans to validate models on unseen engineering cases and secure a first paid industrial pilot. The prize will help the team continue developing its industrial modeling software and testing it with customers.
Second Place, $50,000: NEXUS, Baylor College of Medicine and Columbia University
Led by Shay Beheshti and Brooke Dirvin
NEXUS is developing AI-assisted clinical-trial matching to help cancer patients and research teams navigate enrollment.
The team, made up of two genetics doctoral students, combined its scientific expertise with customer discovery and AI-enabled development. Conversations with clinicians and research staff helped the founders refine their approach around real-world needs. During the challenge, they reported securing a paid pilot commitment. The prize will help advance their clinical-trial matching platform and pilot work.
Third Place, $25,000: Lobe, Texas A&M University
Led by Shubh Bhakta and Tiernan Lindauer
Lobe provides AI tools to help senior-living communities respond to inquiries and follow up with prospective residents.
The team of an undergraduate business student and an undergraduate engineering student advanced after translating customer discovery into paying contracts, reaching four paying communities across two operator relationships. The prize will help the team reach more senior-living operators and build on its early customer relationships.
Looking Ahead
Mays plans to open the 2027 AI Venture Velocity Challenge in spring 2027. A forthcoming report will look beyond the winners to explore how students from different disciplines used AI to test assumptions, learn from real customers, and turn evidence into venture progress. It will also examine what their experiences suggest about how universities can help students combine human judgment with AI to tackle real-world problems.
“Business schools have been doing pitch competitions for a long time. What Mays wanted to do was carve out a unique space,” said Dr. Nate Y. Sharp, dean of Mays Business School. “AI has changed the process of entrepreneurship on a fundamental level. Students from all over the country recognized that we were offering something different, and they chose to be a part of it. We are proud that Mays is seen as a place where student entrepreneurs across the nation can choose to engage in our competitions, learn critical lessons while building a real company, and then come here to share what they tested, what they learned, and how that changed what they built. We look forward to using what we learned this year to build an even more impactful competition in 2027.”
Explore the 2026 AI Venture Velocity Challenge results and finalist ventures, and learn more about AI at Mays Business School.
The AI Venture Velocity Challenge is made possible by the generous support of Deloitte, the Deloitte Foundation, and Midjourney.
About Mays Business School
Texas A&M University’s Mays Business School is building a better future through business. By providing leadership-centered, experiential education to more than 8,000 undergraduate, master’s, and doctoral students in accounting, finance, management, management information systems, marketing, and supply chain management, Mays consistently ranks among the top public business schools nationally for its programs and faculty research. Our mission is developing leaders of character who make a positive difference in the communities where they live, work, and serve.
