How Microsoft Makes Capacity Planning a Dynamic Strategy

AI can sound perfect in the boardroom but fall apart on the factory floor, a gap Dayan Rodriguez understands better than most. Long before he was the Corporate Vice President of Manufacturing and Mobility at Microsoft, he was an engineer writing code directly for the shop floor.

This frontline perspective followed him through leadership roles at Rockwell Automation and Honeywell, where he learned that plant managers value safety, quality and measurable ROI over high-level buzzwords.

Today, Rodriguez leverages that operational grit to bridge the divide between Microsoft’s digital solutions and the gritty realities of industrial production.

Dayan shares his expertise with Manufacturing Digital.

See the full story in the May 2026 edition of Manufacturing Digital.

How does capacity become an active lever rather than just a plan?  

Capacity becomes a lever when it’s dynamic. You need real-time visibility into demand, materials, labour and equipment, along with the ability to act quickly.

When teams can rebalance production or shift resources in hours instead of waiting for the next cycle, capacity becomes strategic. AI helps surface constraints and trade-offs so leaders can move with confidence. 

What does an intelligent system see that a standard process misses? 

Standard processes show what should be happening. On the other hand, intelligent systems show what is happening. 

I’ve seen small stoppages, just a few minutes at a time, add up to hours of lost output each week. AI catches patterns like that.

It connects changeovers, quality drift and supplier variability across silos. That’s where the real opportunity has become, understanding cause and effect across the system instead of reacting to yesterday’s metrics. That’s the difference maker. 

Where is AI’s biggest return on investment? 

ROI shows up where variability is high and decisions matter every day: downtime, scrap, schedule adherence, inventory and even safety. 

When I carried a P&L, I focused on operational performance and margin. AI earns trust when it improves OEE, reduces unplanned downtime or prevents a costly failure within months. The key is embedding it into maintenance, scheduling and quality workflows.

If it lives only in a dashboard, it won’t move the needle. It needs to be built into the culture, progressing with team upskilling, with a consistent vision from leadership. 

How do you design AI that fits how people actually work? 

Start by spending time with the person doing the job. I’ve stood next to maintenance technicians who don’t want another screen. They want clarity. This is key when designing AI. 

Design it to answer a simple question about what to do next and why. Build it into the systems teams already use. Add clear guardrails and transparency.

When AI reduces friction instead of adding complexity, adoption follows. People need to benefit, and human ambition can be the unlock. 

How much action can AI agents take across supply chains? 

With the right controls, quite a bit. 

Today, agents are effective at monitoring inventory and supplier commitments. More and more we’re seeing logistics events, capacity signals, then flagging issues and triggering workflows. The agents are reasoning effectively.

Over time, they can automate low-risk decisions within defined policy limits while escalating complex ones. Speed matters, but accountability matters more and humans are central to leading the success with agents and supply chains. 

What can manufacturers do to get the best advantages out of AI? 

Focus on a small number of high-impact use cases tied to throughput, quality, safety or working capital. Measure them clearly. 

I’ve seen organisations launch too many use cases at once and stall or fail. The companies that win prove value quickly, build confidence and scale from there. 

Data is such a critical element. Invest in your data estate, but invest just as much in people and change management. AI becomes powerful when it’s embedded in daily operations fuelled by the best data.

We’re seeing that more and more across our partner ecosystem and customer AI transformation at Microsoft. 

How do you see capacity planning changing in the next year? 

Capacity planning is moving closer to execution. Instead of static forecasts, teams will rely on live production data, supplier updates and demand shifts to run frequent what-if scenarios. 

AI will handle much of the heavy analysis so planners can focus on decisions and trade-offs. Planning will reflect real bottlenecks, changeovers, skills constraints and maintenance windows.

It becomes faster, more grounded in reality and more tightly connected to financial performance.

See the full story in the May 2026 edition of Manufacturing Digital.