Rob grew up watching his father go to work in a factory every day. That shaped everything - a deep respect for people who build things with their hands and the understanding that hard work is not something you romanticize from a distance. It is something you live around.
After a career in investment banking and strategy consulting, and a Yale MBA, the biggest lesson was not about finance or frameworks. It was that the last bastion of real, honest hard work is the people who work with their hands. The built environment - the people who pour concrete, set steel, and ship the structures our society depends on.
Ody built Sensytec, developing patents that advanced quality control in the precast industry. Real technology, solving real problems on real plant floors. We first met in college, and years later reunited to take Sensytec into commercialization and market adoption.
What we kept hearing
Throughout the Sensytec journey, we were in and out of precast plants constantly. Talking to owners, plant managers, foremen, quality teams. And the same pain point came up over and over: the software is terrible.
Not just imperfect. Terrible. Plants running on spreadsheets held together by one person who knows where everything is. Others stuck on legacy systems like Titan or Concrete Vision that have not meaningfully evolved in years. Some trying to make SAP work for an industry it was never designed for.
Every conversation ended the same way. Someone would say: "If you ever build something better, call me."
Why nobody has fixed this
The answer is simple: tech people do not understand this industry. Every project is bespoke. Every company runs a little different. The guy building parking garages in Texas has a completely different operation than the guy making bridge beams in Ohio, even though they are both "precast manufacturers."
You cannot parachute in from Silicon Valley, spend a few months studying the space, and build something that works. You have to know the difference between a double-tee and a spandrel. You have to understand why a bed is a scheduling constraint and not just a location. You have to know that PCI audits can shut you down and that QC is not a checkbox - it is the backbone of the operation.
This is why generic ERPs fail here. They are built by people who think a piece mark is a SKU.
Building from scratch
We did not try to adapt an existing ERP. We built CastOS from the domain up. The data model understands precast - piece marks, pour schedules, bed utilization, mix designs, cure times, strand stressing. It is bilingual from day one because half the workforce in North American plants speaks Spanish as their first language.
We built it for the foreman on the floor with a tablet, not just the office manager at a desktop.
Where AI actually helps
There is a lot of noise about AI in manufacturing. Most of it is impractical or solving problems that do not exist on a real plant floor. We are focused on the places where AI saves hours of manual work without requiring anyone to trust a black box.
Upload a drawing package and AI extracts the piece marks, quantities, and dimensions. It drafts the estimate from historical data. It flags QC checks that are due. The human is always in the loop. AI goes first, your team corrects.
Why us
We are not tech people who discovered construction. We have spent our careers in the built environment. Rob's entire career has been about helping this industry solve its largest problems. Ody literally invented technology that changed how precast plants do quality control.
We are building CastOS because we know what this industry needs. And we believe we are the right people to do it.