Key Takeaways

  • The traditional SaaS founding pair, splitting commercial duties and software engineering, breaks down in physical hardware and deep tech ventures.
  • Physical design changes trigger immediate ripple effects across subsystems; lengthening a robotic arm by one inch alters lifting payload, mechanical stresses, software controls, and computer vision models.
  • Engineering teams in deep tech span distant fields, forcing founders to manage direct handoffs between nuclear physicists, mechanical engineers, electrical technicians, and software developers.
  • Antares founder Jordan illustrates the target profile by combining nuclear physics knowledge, venture-backed development speed, and federal procurement know-how in one leader.
  • Unlike cloud software that scales instantly on Amazon Web Services, physical systems face immutable lead times, making operating rhythm and multi-disciplinary fluency the primary filter for capital risk.

The Software Split Fails in Physical Systems

For two decades, early-stage venture capital relied on a clean division of labor. One founder handled product and code; the other handled sales, fundraising, and hiring. Leo Polovets, General Partner at Humba Ventures and co-founder of Susa Ventures, points out that this structure collapses once a company enters the physical world.

In pure software, product modifications rarely break the underlying infrastructure. In physical hardware, a minor customer request cascades through every layer of the machine. As Polovets explains, “on the hardware side, if you know let's say you make robot arms and some customers like hey can you make the robot arm an inch longer, right? Like it has all these ripple effects like you have to change the software, you have to change maybe like the computer vision that the robot's trained on. Now it changes like how much the robot can lift because the arm is longer.”

A non-technical executive cannot assess whether a commercial commitment introduces a six-month redesign. At the same time, an isolated engineer cannot evaluate whether that mechanical change violates a client's operational constraints or triggers a regulatory review. The commercial and technical boundaries must sit inside the same brain.

Managing the Cross-Discipline Communication Tax

Software architectures favor modularity. An engineer can push an update to an isolated microservice without knowing how the payment gateway was coded. Deep tech projects operate under tight physical interdependencies. Polovets notes that instead of managing a uniform team of software coders, deep tech founders run organizations where “we have like a nuclear physicist and a mechanical engineer, an electrical engineer, and a software engineer. And each of those people has to communicate with the other ones on what's going on on their side or their bottlenecks or they need changes for somebody else.”

Polovets cites Jordan, founder of microreactor developer Antares, as the archetype for this environment. Antares does not separate policy navigation, reactor physics, and engineering pace into distinct silos. Jordan combines technical physics comprehension with startup speed and defense procurement background. When these domains collide during development, decisions happen without cross-departmental translation delays.

The penalty for miscommunication in physical systems is time. Code scales on demand. Polovets contrasts that ease with the reality of atoms: “if you're building a software business like you build it you put on AWS and it like kind of scales automatically for you on on the deep tech side there's a lot of interdependencies and a lot of them have a lot like long time constraints.” Waiting for custom actuators, specialized alloys, or government permits means that design errors compound into years of stalled runway.

Why It Matters

As artificial intelligence reduces the barriers to producing pure software, defensibility is migrating toward physical infrastructure, power generation, data center hardware, and specialized semiconductors. Underwriting these companies requires changing how investment committees evaluate management teams. The generalist MBA paired with a technical lead is becoming an operational liability in physical tech, shifting early valuations and capital allocations toward operators who possess native fluency across physics, hardware supply chains, and government channels.