About this Presentation
Jack Warchalowski and Duncan Patrick examine a practical question: how can organizations ensure that AI adoption produces system-level business value rather than simply making isolated tasks faster? They frame AI through a technology-and-productivity lens drawn from TOC: understand the technology’s power, identify the limitations it removes, recognize the rules and behaviors built around those old limitations, and then change those rules to capture the benefit. Demand Driven Replenishment is used as a comparison: its value depends on moving away from forecast dependence, managing at SKU level, following buffer signals, and relying on a dynamic feedback loop. Applying the same logic to AI, they ask whether data is clean enough, whether reviews and checklists are still necessary, and whether AI may simply accelerate work that should not be done.
The presenters connect this to system focus. AI should be applied where it improves flow through the system constraint, while measurements, rules, and behaviors are adjusted to prevent local optimization. Their own software-development process illustrates the risk. Developers reported dramatic time savings from AI-assisted coding and review, yet the measured flow rate improved by only about 5% while cycle time and WIP stayed essentially constant. The diagnosis was that saved coding time was being absorbed by additional review activity; without changed rules, local efficiency did not translate into materially higher flow.
They then demonstrate Montera AI within the Roadrunner Rx inventory-management service. The AI is used to investigate overstock, stockouts, over-ordering, planner compliance, buffer sizing, and what-if questions, and to support data cleansing, initial and ongoing buffer sizing, error analysis, and capacity. The deck also records lessons from creating an AI agent: it must be taught, starts without memory, can make up answers, and consumes increasing token costs. The closing implication is that AI does not diminish the value of TOC. The presentation argues that organizations still need focus, adequate data, subordination, and explicit changes to rules, behaviors, and measurements to turn faster technology into business results.
What Will You Learn
To help you get the most value from this session, we’ve highlighted a few key points. These takeaways capture the main ideas and practical insights from the presentation, making it easier for you to review, reflect, and apply what you’ve learned.
Why faster AI-enabled tasks do not automatically create business value: the rules, behaviors, measurements, and assumptions built around the old limitation must also change.
Why TOC focus matters for AI adoption: apply AI where it improves flow through the system constraint rather than trying to optimize every activity.
What Montera learned from AI-assisted software development: despite large local time savings, flow rate improved only about 5% while cycle time and WIP stayed constant because saved coding time was absorbed by more review.
How AI can support demand-driven inventory management through data cleansing, buffer sizing and resizing, error analysis, what-if scenarios, and investigation of overstock and stockout causes — while still requiring trustworthy data and disciplined subordination.
Instructor(s)
Jack Warchalowski
Jack Warchalowski is the CEO of Montera. Montera is a global software and consulting company that helps manufacturers solve persistent operations and supply chain problems to accelerate productivity, revenue and profitability.. Jack helps organizations enhance their profitability and competitive position through the implementation of strategic improvement initiatives driven by Roadrunner software. Jack is a Certified Management Consultant and a Professional Engineer registered in Ontario. He holds an MBA degree from the Wilfrid Laurier University and a Bachelor of Applied Science in Mechanical Engineering from the University of Waterloo in Waterloo, Ontario. In addition, Jack is certified by the TOCICO in all aspects of TOC.
Duncan Patrick
Duncan Patrick is Executive Vice President of Montera Inc. Montera specializes in software and consulting that helps clients solve problems in Operations and the Supply Chain. Duncan’s business consulting career is focused on working with clients to assist them solve problems in operations and the supply chain related to lead times, the forecast, inventory, capacity, on time delivery, product development speed and market focus. Duncan holds an MBA degree from the Richard Ivey School of Business, Western University and a Bachelor of Commerce degree from The University of Calgary. Duncan is certified by the Theory of Constraints International Certification Organization in all aspects of TOC. In addition, Duncan is a Certified Management Consultant.