Sonic Modules · Prefab Modular Building System

Four benchmarked companies map a clear spectrum: fully vertical (BoKlok, Lindbäcks), asset-light (Sustainer), pure software (AplusV). The capital-heavy players carry the greatest risk: BoKlok's Swedish operations closed in 2025 and its Gullringen factory was sold. The lighter models move differently: Sustainer licenses per m² with near-minimal burn, and AplusV bootstrapped from a TU Delft thesis to 25+ clients on effectively zero outside funding. A factory has to run 24/7 to be profitable, and companies that own one tend to become so absorbed in production that design, business model and flexibility quietly starve. That is what pushed Sonic Modules to software-first: the advantage sits in the design system, the network, and the data each project generates.

• 01

  Every rule was a hypothesis, not a decision

  Every rule locked into the system, whether a grid dimension, an invariant or an export format, was framed as an assumption and shown to builders, engineers and prefab specialists. Where it held it was kept; where it didn't, that part of the app was reworked. That loop, not any single technical fix, is what shaped the final eight phases.

• 02

  The choice of building system is not a detail

  The assumption was that picking a building system is a limited, largely technical choice. It isn't. The system drives overall project cost, and standardization, assembly time and detailing can all be tuned to the client. That reframed the technology phase from a settings screen into a genuine decision point.

• 03

  Hardware would have killed the flexibility

  The starting hypothesis was that owning a production line gives the most control and the best path to scale. The interviews reversed it: strategy has to match capital, and at bootstrap stage hardware removes exactly the flexibility an early venture needs to pivot.

• 04

  The moat is the data, not just the product

  Advantage was assumed to come from the product and the network. It also comes from what accumulates: build → learn → create data → repeat. Every delivered building makes the next one faster to win and deliver, which is why the go-to-market diagram closes with a flywheel rather than an arrow.

• 05

  Slowing down at the start saves time later

  My instinct was to move fast between stages and validate by building. Every stage of a venture needs its own pace, and resisting the urge to build during problem discovery, talking to as many people as possible instead, is what saves time downstream. A genuinely annoying lesson for anyone who would rather be writing code.

The software generates a coherent building system on screen; whether the studs, beams and connections survive real engineering and manufacturing without manual rework is unproven. Next is a structural engineer's review and one test module fabricated from a generated drawing set, because assembly tolerances can only be tested by building them. The process has not yet run end to end with a first client, and the willingness to pay for software without Sonic Modules controlling manufacturing remains the assumption everything rests on.

The venture closed with a pitch and jury Q&A at BK-Launch, TU Delft's entrepreneurship programme at the Faculty of Architecture, in front of founders, investors and practitioners. The photographs below are from that demo day, shot and published by BK-Launch. It is the room where a 64-page business plan had to survive being reduced to five minutes.