We’ve Been Quietly Building “Intelligent Hardware” for Years—Now We’re Saying It Out Loud
- Dr. Macarena Cataldo
- Apr 22
- 2 min read

At Viridis, we've been using Intelligent Hardware - aka hardware powered by AI, data modeling and advanced simulations - before it was cool, and definitely before it was on every pitch deck and conference panel. For the past four years, we've relied on tools like computational fluid dynamics (CFD), machine learning and quantum chemistry to do what we do best: move fast, build smart, and avoid throwing money down the drain.
Why didn’t we talk about it before? Simple: To us, these tools aren’t a marketing gimmick, they’re just a necessary part of our everyday work. If you're serious about building high-performance, next-gen hardware to solve real-world problems, you have to ditch the “build it and pray” method. We use modeling and predictive simulations because blowing $70K on each prototype iteration is, quite frankly, ridiculous.
But it wasn’t until a conversation with the technical advisor of an investment group, someone genuinely curious about how we advanced so fast, that I realized we should probably start using our outside voice on this. When I explained how we combine machine learning with deep chemistry knowledge and advanced design modeling, they were like, “Wait, why aren’t you shouting this from the rooftops?” .
So fine. Here we are. Shouting:
At Viridis, we use:
⚙️ CFD (Computational Fluid Dynamics):
To optimize how fluids move inside our electroxidation reactors - because better flow = better mass transport = better performance. Period.
⚙️ Machine Learning + Statistical Design of Experiments:
To cut down on the number of physical tests, identify key variables, and get answers faster—with fewer chemicals, less waste, and more clarity.
⚙️ Quantum Chemistry:
To screen new materials and predict surface reactivity before wasting time and resources making them in the lab.
⚙️ AI-Powered Predictive Models:
To monitor and control our systems in real-time, helping us (and our clients) spot issues before they become expensive problems.
And the economic benefits? Let’s break it down:
💰 Lower R&D costs.
We’re not running blind experiments or building version 12.3 of a reactor that could’ve been optimized digitally.
💰 Faster time to market.
Modeling lets us move from idea to results without the months-long lag.
💰 Better system design.
An intelligent, data-driven design process means our clients get more efficient, economical solutions right out of the gate.
💰 Less waste.
Fewer failed prototypes, fewer bad material choices, fewer headaches.
But here’s the most important part: all of this only works because our technical team is absolute 🔥. You can throw all the machine learning you want at a problem, but if you don’t know what you're doing, or don’t understand the underlying science, you’re just generating pretty graphs that mean nothing. Our AI tools are powerful because they’re built with (and not as a replacement for) deep engineering and scientific expertise.
Oh and in case you are wondering: just like our groundbreaking technology, this blog post was created using the power of the human brain, with the gentle support of AI 🧠🤝🤖
Comments