2025 AIChE Annual Meeting

(492d) From Plant to Platform: Applying Chemical Engineering Tools to Build an AI Startup

Author

At 13, a future chemical engineer stepped inside an internet startup’s headquarters in California. That summer revealed a different kind of engineering—where founders chased possibility not because they had permission, but because they knew how to build. That early exposure, paired with a passion for sustainability and systems, sparked a career that would later blend classical chemical engineering with agile software thinking. The journey explored in this paper follows a chemical engineer who blended startup thinking with classical process improvement to build a new kind of tool for industrial teams. Many early-career engineers face a Catch-22: without a PhD or millions in venture capital, the typical pathways to entrepreneurship in chemical engineering—novel membranes, catalysts, or CRISPR applications—can seem inaccessible. Meanwhile, software founders build fast, fail faster, and iterate with conviction alone. Process engineers carry a powerful toolkit: Designed Experiments (DOEs), control systems insight, and data modeling. What happens when those tools are aimed not just at plants, but at platforms?

This paper shares how the statistical methods used in process optimization—often viewed as back-end support—can become front-end engines of innovation. Plant-wide DOEs, neural-network efficiency modeling, and the analysis of moisture-ash-turbine interactions became stepping stones toward building an AI-first root cause analysis (RCA) platform. One that doesn't chase a single root cause but supports team collaboration, visualizes data patterns, and speeds up organizational learning. The transition from fieldwork to founder wasn't solitary. Along the way, mentorship from world-class experts—including a few PhD chemical engineers, entrepreneurs, and a Master Black Belt with a passion for statistics—played a critical role. With their guidance, it became clear that loving data wasn’t about spreadsheets; it was about empowering teams to see clearly and act faster.

This session will explore: how fail-fast, lean startup methodology can be reframed using chemical engineering principles, why team dynamics and real-time insights matter more than theoretical root causes, how combining DOE rigor with modern ML accelerates decision-making, and what chemical engineers can learn from software—and vice versa.

Building a company this way is a risk, but one rich with opportunity. This is not a story of overnight success. It's a blueprint for turning field experience into scalable impact. For engineers curious about entrepreneurship, innovation, or simply getting better at solving real problems faster, this paper offers both technical insight and an honest reflection on what it takes to start building before waiting for permission.