Arkansas will be the latest state in the United States to adopt Dynamic Infrastructure’s technology for infrastructure analysis, as governments from 13 US states already use the platform, according to Saar Dickman, CEO and founder of the company. The startup’s pitch is simple enough: sell authorities a machine-assisted way to decide which crumbling systems get attention first, while the people living under those systems wait for the verdict.
Who Gets to Decide
Dickman said the platform helps civil engineers process large amounts of data to determine which infrastructure needs to be prioritized. In a world where infrastructure is aging and budgets are limited, he said, the system enables authorities and state transportation agencies to gain clear visibility into the condition of their assets and manage them efficiently and proactively. That is the language of administration from above: assets, priorities, efficiency, proactive management. The people who depend on bridges, roads and other systems do not appear in that vocabulary except as the ones who absorb the consequences when the apparatus decides what matters.
Dynamic Infrastructure was founded in 2019 by Dickman and Amichai Cohen. The company reported a 100% contract renewal in 2025 and said it plans to expand to the Australian and European markets, with the objective of providing every public or county engineering and maintenance department with an AI-based “virtual engineer” that works alongside professional teams and delivers unprecedented force multiplication. The company reported $1 million in revenue in 2025 and projected that it could triple that in the coming year.
How the System Works
Dickman said the company’s approach to liability is to use human checkpoints in the system because AI is not yet developed enough to be used in a fully autonomous way. He said the information processed is collected by a certified inspection engineer or a certified contractor, who is paid for the information, and that from that moment on the infrastructure owner is liable because he paid for the inspection service and is supposed to know what is happening once the inspector provides the results. He said the platform only helps in going through that information.
He also said the system is not completely run by AI and that civil engineers revise the work at points in the analysis, adding an extra layer of reassurance to the final result. He said the company does not aim to replace civil engineers, but to become a helping hand when processing massive amounts of data, adding that artificial intelligence cannot fully replace engineers and can only support them.
That “helping hand” is built around a chain of certified labor and liability, with the infrastructure owner left holding the bag once the paid inspection is delivered. The platform does not remove hierarchy; it streamlines it. It turns the old bottleneck of human review into a faster pipeline for the same top-down decision-making.
Training the Machine, Selling the Fix
Dickman said one challenge in developing the system was explaining the difference between modern structures, which have 30 to 40 years of data from their construction to the present, and antique structures hundreds of years old. He said that once the system is trained and the company’s unique IP is used, it goes very smoothly. He said the company had to train the difference between a brick falling from a modern brick and one falling from a medieval arch built 400 years ago.
Dickman said the advantage was that the system was developed with a team of civil engineers, not just programmers, so they knew what the AI needed to learn, where to find mistakes and how to correct them. He gave an example from the early development of the system, saying that a photo from one of the company’s clients in Greece showed a red-haired woman standing on a bridge, and the system identified the woman as rust on the bridge. He said that does not happen today, but it remains both a learning experience and a funny story.
The company’s sales pitch is wrapped in the usual techno-managerial promise: more data, more speed, more “visibility,” more control for the institutions already tasked with managing public infrastructure under conditions of scarcity. The platform is presented as a solution for state transportation agencies and county departments, not as a way for ordinary people to decide what gets repaired, when, and for whom. The result is a cleaner interface for the same old arrangement, where public need is filtered through certified experts, paid inspections and institutional liability.
For now, the company says its system is already embedded in state-level infrastructure analysis across 13 US states, with Arkansas next in line. The machine may be called a virtual engineer, but the real power remains where it always was: with the authorities, the owners and the departments that get to define what counts as priority, what counts as risk and what gets left to rot.