What happens when production planning gets too big for spreadsheets

Matthew Sabourin built his own production scheduling system before he ever thought about buying one.

He pulled it together with his intuition, spreadsheets, and years spent waiting for beer to ferment at Nonsuch Brewing Co., a majority Indigenous-owned craft brewery in Winnipeg’s Exchange District. It had an intelligence to it, but the whole system lived in Sabourin’s head, and when the brewery needed him elsewhere, it seized up.

“It got so onerous to me, we had to shelve the tool,” says Sabourin, president and co-founder of Nonsuch. “We just stopped using it, because I needed to be other places.”

Most companies have one. The spreadsheet that started out as a quick fix that became a system, which then turned into a small hostage situation.

An industry friend saw what Sabourin had built, and told him he needed to talk to Cameron Bergen, founder and CEO of Mode40, a Steinbach, Manitoba-based enterprise data intelligence firm.

The two joined a Canadian Food Innovation Network webinar on May 28, moderated by Digital Journal, to walk through how they’re solving it. 

Nonsuch isn’t alone in this. In manufacturing. When operational knowledge lives in one person, the cost shows up on the production floor as capacity you didn’t know you were losing.

Capital expenditures across Canada’s food and beverage sector declined 5.3% in 2025, according to Farm Credit Canada (FCC), and early indicators suggest investment could weaken further in 2026. The sector is also short 50,000 workers, according to the Canadian Agricultural Human Resource Council. 

For most of the sector’s 11,000-plus businesses, that starts with the operations they already have. Unfortunately, some of it’s hiding inside a spreadsheet named “Final_FINAL_revised_v7.” 

If you’re lucky, it’s marked with a date or someone’s initials. 

Nonsuch and Mode40’s partnership is a working example of what it takes to bring AI onto a small manufacturer’s floor, including the messy data, the buy-in, and the slow build.

Small operations carry billion-dollar complexity

Beer sounds simple when you’re the one ordering it at the bar. But behind the scenes, the customer wants it tomorrow, the tank needs three more days, and the production spreadsheet starts feeling more like you’re painting while blindfolded.

Bergen has solved this kind of problem at larger facilities. Nonsuch packs it into a fraction of the space.

“It’s probably one of the hardest production scheduling calculations that we’ve seen, even across multibillion-dollar enterprises,” he says.

You can’t tell a batch of beer when to finish fermenting. One extra day in the tank cascades through the entire plan.

“Why not just make big batches of everything and leave this on the shelf forever, right?” Bergen says. “Because quality is too important to what they do.”

The underlying problem is familiar to any operation with unpredictable outputs and customer demands that arrive faster than production can respond. 

At Nonsuch, the variable is biology. We’re talking about yeast and the environment required to do its job. It’s done when it’s done. 

Elsewhere it might be supply chain delays, machine variability, or shifting customer specifications. The details change, but the headache is familiar.

Mode40’s system creates a digital model of the brewery’s production process, then uses Nonsuch’s own rules around quality, capacity, and customer commitments to help schedule what should be brewed, packaged, and prioritized.

Before Mode40, Sabourin managed all of it across multiple manually updated spreadsheets. One beer might need to be packaged into different keg sizes and can formats, while multiple batches sit at different stages of fermentation, and orders arrive the same day they need to ship.

“I built a spreadsheet that is populated manually, updated again and again and again manually, [and] prone to human mistake,” says Sabourin.

Unfortunately, every missed production window compounds. 

“Once that opportunity is gone, you can’t get it back,” he adds. “It’s gone forever.”

One of the less glamorous problems was getting the data out in the first place. Sabourin says Nonsuch’s enterprise resource planning (ERP) didn’t have an export button, so Mode40’s team pulled the information from the webpage code instead. 

It doesn’t get more small manufacturer-coded than “the data exists, technically, but good luck retrieving it.”

AI can’t fix what it can’t see

Bergen starts every consultation on site, because what a manufacturer describes on a call and what he finds on the floor are almost always different problems.

“If you haven’t seen it, don’t try to scope it,” he says.

Mode40 builds a digital model of the facility first, then layers the operator’s decision rules on top. When two orders need the same tank on the same day, what wins? The bigger customer, the tighter margin, the beer that can’t wait?

“You’re really going to answer them,” Bergen says. “Those sets of rules become the foundational barriers to the way the system works.”

Without that foundation, even advanced AI will produce outputs that look right on screen and fall apart on the floor. 

This goes well beyond food and beverage. Implementing AI before mapping the operation is how companies end up with a confident answer to the wrong problem.

The Nonsuch-Mode40 partnership is ongoing, with sales forecasting work still in progress. Bergen says one of the early opportunities is making capacity issues visible before they become missed orders.

He says the goal is to spot conflicts early enough that manufacturers can adjust production or call clients before a missed order becomes a relationship problem.

“You’ve mitigated the risk ahead of time without ticking off a client,” says Bergen.

Nobody wants another screen to babysit

Sabourin came into the implementation process more technically confident than most, having built his own platform in a previous business. Even so, his employees had questions about the six different systems they had stitched together. 

“Without full buy-in, you’re going to be struggling uphill the entire time,” he says. “So you need everybody to believe that this is the right solution, so that they can be a partner in the implementation rather than resistance.”

Fair enough. Most workers have met at least one system that was supposed to save time and somehow created a new part-time job.

Mode40 runs facilitation sessions before implementation starts to surface concerns early, everything from data privacy to scars from legacy systems that promised more than they delivered. Bergen says most of those fears come from how technology used to be sold. 

“The more complicated it is in that fashion, [the more] likely they’re selling you on the idea,” he says. 

Modern tools should be harder to explain than they are to use. Bergen’s read is that when a platform overwhelms you with screens and features, you’re paying for the complexity, not the results.

For Sabourin, the payoff is the physical plant. Nonsuch has a fixed number of tanks and a fixed amount of floor space, and the goal is squeezing more out of both. 

“The more we can get out of it, the better,” he says.

A new FCC report found that hitting 3% annual GDP growth in Canadian food and beverage manufacturing over the next decade could add $40 billion to the national economy and create 217,000 jobs. 

For the SMEs that make up most of the sector, that kind of growth starts the way it did at Nonsuch, by getting more out of what’s already on the floor.

Final shots

• Many manufacturers are running on institutional knowledge trapped in one person and forecasting spread across disconnected systems.
• If an AI platform can’t map how your operation works before it starts automating, the outputs will likely be wrong, despite sounding confident.
• When evaluating new tools, Bergen’s test is whether the architecture lets you start with one problem and grow from there.

Watch the full webinar, moderated by Digital Journal, below: