Georgia-Pacific didn’t treat AI as a science project or a luxury add-on while researching solutions. They made it core to operations, scaling intelligent decisions across an enterprise as massive as their manufacturing footprint.
Georgia-Pacific manufactures products used every day, like bath tissue, boxes and lumber, across more than 150 sites worldwide, spanning manufacturing, logistics and office locations. From household brands such as Brawny® paper towels and Dixie® cups to building products like DensElement® Barrier System, Georgia-Pacific operates at a scale where even momentary downtime can translate into millions in lost revenue.
To mitigate risk, enhance safety and optimize efficiency, they turned to SAS® Viya® and SAS® Intelligent Decisioning.
Cracking the recausticizing code: How Georgia-Pacific stabilizes centuries-old process with AIBut first, your data must be AI-ready
Before Georgia-Pacific could make smart decisions in real time, it had to wrestle with disparate, siloed data. Like many enterprises, Georgia-Pacific had grown through acquisition and blood, sweat and tears. Launching AI wasn’t just proprietary algorithms and flashy headline-grabbing projects. It was going to be down-to-earth, ‘data 101.’ They needed access to trustworthy, usable data.
"We had systems from the ‘80s, ‘90s and 2000s,” said Adam Frowein, Senior Director of Connected Systems. “The data was all in different formats, different systems and sometimes with different levels of reliability."
From vibration sensors and temperature logs to computer vision streams and legacy system exports, the challenge wasn’t a lack of data – it was the volume and variety. Different teams owned different sources. Some information was siloed. Some processes hadn’t been digitized at all. SAS Viya’s architecture allowed Georgia-Pacific to integrate and standardize data from across facilities. This enabled Georgia-Pacific to transform several information sources into a single, decision-ready system.
Data + AI automation = fast decisions at scale
Modern manufacturing isn’t just pressing pulp into paper cups. It’s a volatile ballet of chemistry, temperature, pressure and timing. Legacy systems once stymied critical decisions due to IT bottlenecks or scattered rules hard-coded in Python scripts. Georgia-Pacific knew that wasn’t cutting it anymore.
"It used to take eight or ten people to manage our rules infrastructure. Now it can be done by one person working part-time," said Frowein. "Technical experts and engineers – not just developers – can build rule sets that govern everything from board alignment to risk mitigation with SAS Intelligent Decisioning."
The clarity that comes from managing the rules infrastructure effectively has ripple effects. Where it once took up to 30 minutes to assess machine variables, Frowein’s team now gets decisions every minute – more data, faster action. In high-stakes manufacturing, that’s the difference between continuous production and costly downtime.
Teams have the flexibility to test ideas and scale the ones with the most impact. Newly trained employees can operate systems with ease. The result? Smoother operations, more consistent pulp quality and greater confidence among newer staff tasked with managing critical mill systems.
It used to take eight or ten people to manage our rules infrastructure. Now it can be done by one person working part-time. Adam Frowein
Executing in real-time at scale
Manish Sinha, Vice President of IT for AI Architecture and Delivery, described the magnitude of what SAS Intelligent Decisioning enables, "For one of Georgia-Pacific’s signature products, we now make 6,000 calls a second, running 40,000 to 50,000 models a minute. That is a billion-fold increase in what we can execute at scale – in real-time."
That scale isn’t just about volume. It’s about autonomy. "Chemical engineers aren’t waiting on IT; they can now write and deploy their own business logic," said Sinha. That self-service model allows for faster adjustments, tighter process control and fewer chances for communication lag to impact operations.
But faster processes and communications don’t mean risk. The solution provides guardrails against risk, allowing team members to use generative AI in critical applications with zero hallucination.
Chemical engineers aren’t waiting on IT; they can now write and deploy their own business logic. Manish Sinha
A real partnership
For all their in-house ingenuity, Georgia-Pacific emphasizes that SAS isn’t just a vendor – it’s a partner. "We call it a ‘cycle of mutual benefit," Bakalar noted. "They’re as invested in our success as we are. They guide, support and co-develop."
The results speak for themselves: nine-figure gains in productivity, yield, and downtime prevention. Billions of data points are processed in real time. A workforce is safer and smarter than ever.
In a world where milliseconds matter, Georgia-Pacific and SAS are proving that the best decisions are the ones you don’t have to wait for.
