Better Data, Better Decisions: How Tuffaloy Closed the Execution Gap on Epicor

The decision: an ERP integration that actually closed the loop

Tuffaloy did not start with MachineMetrics. The team evaluated several vendors and got far along with a different one before backing out.

“We knew before going into this that our ERP integration was going to be super important. We didn’t really understand all of the details about how it was going to work, whether it was going to be unidirectional, bidirectional, and how that data was going to flow back and forth. But we did understand that we wanted the real-time machine data connected to the Epicor data.”

Jared Talladay, Operations Manager, Tuffaloy

Two factors drove the switch:

• Open, accessible API documentation. “The one thing I noticed before we even started a trial with MachineMetrics is all of their API documentation is online. It was accessible before we were even a customer. That was huge.”
• Hands-on integration support. Paul Richards from MachineMetrics worked on-site at Tuffaloy to validate the connector before rollout.

The rollout: pilot first, one machinist first

Tuffaloy ran the integration in an Epicor pilot environment for several months before going live. Talladay audited every labor transaction.

Production rollout started with a single machinist, intentionally chosen.

“We restricted it to one machinist, our more senior machinist who typically struggles a little bit more than others as far as computer use and technology. We figured if we can walk him through it and get him comfortable, it’ll be a lot easier to get buy-in from everybody else.”

Jared Talladay, Operations Manager, Tuffaloy

Once the bugs were ironed out, the rollout expanded across the 14 connected lathes.

What changed: data became information

Cycle times that match reality

Real cycle time data now flows from every connected machine. Engineering pulls actual times when quoting similar parts. Standards in Epicor can be updated based on what the machine actually does, not what someone wrote down years ago.

“It’s allowed us to really hone in on cycle times that have gotten worse over time and be able to address that out of the machine in real time, because MachineMetrics is putting that data in front of our face saying, this is running slower than expected.”

Jared Talladay, Operations Manager, Tuffaloy

Customer service that can answer the question

The customer service team now runs a live MachineMetrics dashboard on a department monitor. When a customer calls asking to expedite an order, customer service can see how many pieces have been completed, how much time is left based on actual part-to-part data, and the current machine state. The estimated time remaining is built from real cycle data, not stale standards. No more chasing the floor for status.

Downtime classified automatically, with a price tag

MachineMetrics automatically tagged downtime tied to manual bar changes on Tuffaloy’s high-volume Swiss cell. The data showed an average of 10 hours per month, per machine, lost to manually loading bars.

That data justified a capital expense. Tuffaloy replaced manual bar feeders with magazine-style automatic loaders across six lathes. The downtime category virtually disappeared, and the cell unlocked lights-out machining for the first time.

The chart that made the case was built with Max AI, the MachineMetrics AI assistant.

“It took 30 seconds to a minute to make this chart.”

Jared Talladay, Operations Manager, Tuffaloy

Auto part counting that fixed inventory

Once operators understood that MachineMetrics was counting every part automatically, the inventory picture changed. Scrap had to be reported as scrap or it counted as good. Operators no longer tracked counts on Post-it notes across six machines.

The result: an 85 percent year-over-year reduction in inventory corrections from 2024 to 2025.

Quality issues raised in real time

A simple NCR request workflow lets operators flag a quality issue from the tablet with a single button. The system captures the machine, operator, and current operation automatically. Someone from the quality team is on the floor within 30 minutes.

That replaces a previous reality where uncertain parts went into a bin no one else knew about, or never got reported at all.

The culture shift that made the numbers possible

The technology was the easier part.

“Getting the machines connected took a day. Seeing real-time data from the machine took a day. But building culture takes a long time.”

Jared Talladay, Operations Manager, Tuffaloy

Talladay framed the rollout around support, not surveillance.

“The question becomes, not whose fault is it, but what support do they need, and why didn’t they get it?”

Jared Talladay, Operations Manager, Tuffaloy

In practice, that meant transparency with operators about what data was being collected and why. It meant treating slowdowns and quality flags as inputs to training and process improvement, not blame. And it meant aligning the floor and the office around the same real-time facts.

Eighteen months in, that alignment is the foundation underneath every other result in this case study.

Where Tuffaloy goes next

• Automated standards updates from cycle data. Today engineering decides when to push an updated standard to Epicor based on machine data. The next step is closing that loop further.
• Smarter scheduling. The team is using MachineMetrics scheduling to coordinate job sequence at the machine. The next phase is scheduling around limiting changeover times across cells.
• Continued culture work. Talladay’s view: “You can’t rush something like this.”

The pattern, beyond Tuffaloy

Tuffaloy’s experience reflects a pattern across Epicor manufacturers. The ERP tells you what should happen. The shop floor knows what is actually happening. The gap between them is where execution lives, and where it gets lost.

MachineMetrics closes that gap by connecting real-time machine and operator data directly to Epicor Kinetic, bidirectionally. Standards get challenged with reality. Downtime gets a category and a cost. Off-shifts get visibility. Floor teams get support, faster.

The Epicor investment is not replaced. It is amplified.