Bonus Q&A Episode: The Future of Manufacturing with Reata and Fulcrum

Sunny:

Welcome to this extended episode of Capacity. We have Grady Cope and Adam Zaouague from Reata Engineering & Machine Works. We’re going to do more of a Q&A session and talk about the future—where manufacturing is going, where technology is going, and where the labor force is going.

Welcome, guys. Adam, you’re 19–20 years old, right?

Adam:

Yeah.

Sunny:

You work at Reata. Welcome. We’re going to spend some time talking about the future. We’ve spent a lot of time talking about your journey, but the next step is exploring where things are going.

Adam:

Thank you. Why don’t you direct us—start lobbing some questions around and we’ll get started. I had a question for you. You talk a lot about manufacturing—small- to medium-sized manufacturers—and how you want to create a community of those. And we don’t want to just have five giant manufacturers dominating everything.

From the software side, do you think there should be some sort of system connecting manufacturers? And do you think Fulcrum could drive that? I know a lot of people have tried—associations, software meant to connect manufacturers, platers, painters. There’s usually one expert who drives a process and everyone else just sends the work to them. It’s not our value add.

I’ve seen software try to bridge those gaps, but it hasn’t really worked. We’ve tried using some of it and it hasn’t led to much. So from your perspective: what could drive that change? More dinners like yesterday? More events? More in-person conversations or conventions?

Sunny:

Internally, when we talk about strategy, we try to make it as simple as possible. We distill it into: we have this vision for a digital network of manufacturing—what needs to exist for that to happen? And how do we shrink the gap to get there?

The dinner last night is part of showing people this is a non-zero-sum system—everyone doing better makes the entire manufacturing network dramatically bigger. If even a small percentage of the market believes that, it can tip the industry.

Second: technologically, there needs to be an operating system that’s networkable. Our insight—which is humbling and encouraging—is that manufacturing is already a network. It’s just analog: fax machines (unfortunately), emails, phone calls, site visits.

Our job isn’t to create the network. That will never work when one already exists. Our job is to digitize the existing network without destroying it.

Third: we need a product where everyone who uses it has more throughput than those who don’t—but it can’t be so rigid that it prevents innovation. Smaller manufacturers generate ideas. Five giant companies won’t produce all the small, unique, low-run parts or come up with as many ideas.

If our software lets everyone’s ideas permeate quickly—and they make enough money to survive—this future can happen.

Adam:

At Reata, something a little unique is we have people who can program. We find things in the backend of Fulcrum that are super useful.

How do you feel about more customization on the ERP side? One of your challenges is you have so many manufacturers doing things differently. Trying to make software that covers everything and makes everyone happy seems impossible. Sometimes we might need to bend our process to fit the system.

But how do you feel about customization? We find API endpoints, build little tools, add AI, create automated reports from webhooks…

How do you feel about customers doing that?

Sunny:

I don’t want you to bend your process to the system—that’s a horror of legacy ERPs and something we’re aggressively moving away from.

Sometimes you have to name your “enemy”—not a person, but a system. The enemy for us is the fact that most ERPs spend 85% of total cost on services—consultants building customizations. That makes systems so rigid that new technology can’t get in.

And when the software changes, consultants have to come back and fix everything. The incentive structure is broken.

In the 1980s, that model was unavoidable. But not now. Software development is faster. Shops have people like you who can code.

Our philosophy is:

• Make it open, so you can take it the last mile yourself.
• Isolate things bound by physics—we can be rigid there.
• Make everything else configurable or customizable.

That’s the strategy.

Adam:

There’s a saying: if you’re serious about software, you should be serious about hardware too. One thing I’ve experienced—even with Fulcrum—is that the software is great, but does the hardware always complement it?

Target is a good example of integrating hardware: everything is driven by real-world action. Scan an item and it updates inventory automatically.

Where do you feel hardware fits into manufacturing? You probably don’t want to force people into a specific hardware ecosystem, but from our perspective, being more rigid might make software talk to operators more cleanly and make training easier.

Sunny:

I’m fine with tight coupling—as long as it’s done like Apple does it.

But adding hardware increases complexity non-linearly. And I’m wary of adding too much complexity to our organization. We want to build a full-stack, end-to-end system from quoting to invoicing really well.

That’s our current limit. Once we have more resources, we may move deeper into hardware.

Grady:

I agree. Fully vertically-integrated companies end up breaking that apart—it’s inefficient. And hardware changes incredibly fast. How do you keep up?

Also, chip shortages recently showed: if you build around your own processor and suddenly can’t get parts, you’re dead in the water.

But I do think Fulcrum could provide best-practice videos—“Here’s how top shops implement Fulcrum.” That would be really helpful.

And thinking about hardware long-term: where are the glasses? AR? The Vision Pro? Training, setup, visualization… that’s going to be huge.

Sunny:

We bought Vision Pros to experiment. They’re amazing—even though the ecosystem is early.

Imagine setting up a tool with AR overlays. Or running robotic arms remotely with a vision system. Those are real possibilities.

But right now, hardware isn’t robust enough. Cameras need to be ~8x better before we can do automated dimensional measurement with an iPad. But once they are? Everyone will be able to measure with just a tablet.

We’re constantly experimenting. Accelerometers, cameras, microphones, computer vision models—we can detect whether someone is reading a drawing, looking at the part, if the machine is vibrating, etc.

Over the next 2–3 years, as hardware improves, we’re going to race to implement these things.

Grady:

It’s not going to get rid of jobs. It’s going to eliminate the tedious work so humans do the interesting work—and do it faster.

AI will surface insights instantly instead of a human digging through data. Fulcrum already gives us an incredible data library to analyze.

Big data plus AI—that’s the future.

Adam:

We’ve been exploring machine learning too. We partnered with MSU Denver—they got a big AI grant for manufacturing. With machine monitoring and Fulcrum’s event logs, we can make really good inferences.

A tech CEO told me: “You can’t do AI in manufacturing yet—you don’t have a time-series transformer.”

I told him, “Watch this.”

Fulcrum’s logs are perfect time-series data for LLMs to analyze. You can dissect a job instantly.

Sunny:

Exactly. Predictive manufacturing is just “next-word prediction” but with events.

With clean actuals, you can predict:

• Will I make money?
• Will I finish on time?
• Will I run out of material?
• Am I drifting toward a failure mode?

Humans feel changes intuitively, but AI can see the seismic shifts in the data.

Grady:

What I want before I’m gone is to stop being the “historian.” I don’t want everything dependent on me. AI removes that.

Sometimes my intuition is right, but it’s full of bias. AI will remove that bias.

Sunny:

We can make a “Grady GPT” if you want.

But yes—actuals matter. Short-run shops sell variability, not consistency. They need to understand the variability of part geometries.

Actuals plus geometry is where the magic is.

Adam:

Exactly—digitizing all of that engineering knowledge, lean principles, and making it accessible.

Sunny:

We automate tedious work, create visibility, and unlock the ability to solve complex problems computers can now handle.

That’s why one-piece flow is powerful: it reduces complexity for humans. But with enough computation, we can solve problems humans can’t.

Grady:

Scheduling is a huge example. Humans can’t run all the what-ifs. But computers can.

A customer wants a part a week earlier—how does that affect everything else?

Right now, I can do it—but I create a tsunami and crush three other customers. AI scheduling will fix that.

Same with CAM—computers will eventually outperform individual programmers.

Sunny:

A lot of CAM innovation is held back by closed ecosystems. Machine builders trying to protect their turf. That’s starting to change. Companies like Amada openly admit they need help on the software side.

Once machines open up, a lot of magic will happen.

Grady:

Yeah—sometimes it’s not even the machine builder, but the control manufacturer. They want to keep it closed.

But newcomers will replace them. The market will force change.

Adam:

And culturally, even thinking differently helps. Like starting an operation before the previous one is finished—continuous flow. In Japan, that’s foreign, but for shops like ours, it’s normal and efficient.

Grady:

Or shipping partials—some shops refuse, but it helps customers keep their lines moving and gets us cash faster.

Sunny:

Exactly. The future is nimble, diversified, short-run manufacturing. Gigafactory-style consolidation happens because coordination is too complex. But with a connected network, Tesla could iterate faster on batteries or components.

Customization is what customers want.

Grady:

We want the hard parts—the parts that keep you up at night. We’re worse at the easy parts because we’re optimized for the hard stuff.

Our programmers sometimes complain, but they grow from it.

Sometimes you have to “sell inside your own company.” If we learn to make a hard part, it opens a whole toolbox.

Sunny:

You’ve been asking questions—how do you feel learning all of this from the ground up?

Adam:

I’ve kind of made a name for myself as the “ideas guy.” The hard part is balancing ideas with execution—I only have so much time.

But Reata is perfect for me. Leadership is open to new ideas and new tech.

The technology is already here. People say “we can’t do that yet”—but from my perspective, we can.

It’s exciting. Manufacturing has so much room to grow and optimize. It’s like playing a giant tycoon game in real life.

Grady:

I’m a boomer engineer; everyone else is Gen Z. There’s a huge gap, but it gets filled in. I’m geeky enough to keep up.

Trying new things is how we move forward. Some of the automation these guys are working on might even become products.

We recently took on a high-pressure cylinder test—12,500 psi. Everyone said to outsource it. But I knew the engineers could figure it out. We had catastrophes, but they learned.

The customer had never seen data captured the way we did it. We iterated and solved it. That’s the fun of manufacturing.

Sunny:

I think your legacy at Reata—with people like Adam—will absolutely come true. Manufacturing is held back more by execution constraints than anything else.

Every bit of work you do enables more cool things to exist.

Thank you both for the conversation.

Grady / Adam:

Thank you. This was fun.