Building Foundations: Hexion’s AI Growth Story

Most companies bolt AI onto old processes. Hexion CEO Michael Lefenfeld broke them first.

In this DigiChem Dialogues conversation, Agilis founder and CEO Jay Bhatia talks with Michael about taking over a company shaped by decades of private-equity ownership and a bankruptcy, then rebuilding it around data and speed. Their conversation covers:

• Why Michael says you have to “call your baby ugly” before you can transform
• His first three moves: a culture that tolerates failure, decisions driven by market intelligence instead of relationships alone, and an innovation function built around “the cafeteria” idea he took from Bell Labs
• Why AI wasn’t the starting point, and what had to be fixed first: data integrity, cybersecurity and IT
• How sensors that read each batch of wood, plus Hexion’s process data, let the company tune a resin to a specific customer’s needs, with savings Michael puts in the tens of millions of dollars a year for the industry
• Why a formulator’s job shifts from running one product to guiding AI-assisted decisions, with a human still in the loop
• Hard impact (dollars) versus soft impact (capability), and why large chemical companies have to be willing to break things
• Why the winners will combine science, data and human expertise

Guest: Michael Lefenfeld, President & CEO, Hexion

Host: Jay Bhatia, Founder & CEO, Agilis

Watch and listen on demand: https://digichemdialogues.agiliscommerce.com (use the Hexion episode page URL here)

Connect with Michael on LinkedIn: https://www.linkedin.com/in/michael-lefenfeld-836a41/

Learn about Agilis: https://www.agiliscommerce.com

Chapters (rounded from the transcript timestamps; please check them against the edited audio)

• 00:00 Welcome and guest introduction
• 03:30 “Call your baby ugly”: why Hexion needed a reset
• 05:00 What Michael inherited: Borden, private equity, bankruptcy
• 07:30 The first three moves
• 10:00 The Bell Labs cafeteria and how innovation really happens
• 12:00 In the office versus remote
• 16:00 Was AI part of the plan from day one?
• 19:00 What had to be in place before AI, and why to break the process first
• 22:30 AI for growth: reading wood and tuning resin per customer
• 27:00 Institutional knowledge, AI and the human in the loop
• 30:30 Audience Q&A: regional and recycled wood
• 33:00 From thousands of SKUs to millions
• 35:00 The biggest success: the people
• 36:30 Measuring AI impact: hard and soft
• 38:30 How large chemical companies should approach AI
• 41:00 How agentic is it really?
• 43:00 Procurement: a penny a pound
• 44:30 A culture that accepts failure
• 46:00 The next three to five years
• 49:00 Michael’s message to the audience
• 51:00 How to reach Michael (Hexion is hiring)