Why we invested
Carbon-fibre structures fail quietly. A ply lands a few millimetres out of tolerance, a void forms during consolidation, and nobody finds out until non-destructive testing rejects the part — days of cure time and a five-figure layup already spent. Across aerospace, defence and energy, scrap and NDT rework routinely consume a fifth of the cost of every composite structure built, and the people who prevent it are a small, ageing population of technicians whose judgement has never been written down.
Curelum is building the operations layer that encodes that judgement. Sensors and vision on the tow-placement head, a model that predicts where a defect is forming rather than certifying one after the fact, and a control loop that corrects the path mid-layup. The output is not a dashboard; it is a part that passes inspection the first time, with a certified strength record attached to it.
We led the seed round in October 2025. What convinced us was not the demo — it was that the two founders had spent years on opposite sides of the same failure. One had signed off on rejected structures. The other had built the inspection systems that rejected them. They agreed, without prompting, on which twenty percent of the problem was worth solving first, and they had already earned pilot access to a production line before raising a dollar.
This is the kind of company our thesis is built for: unglamorous, physically constrained, and defensible precisely because the data is impossible to acquire from a laptop.
We were a factory-floor company asking software investors for money. Connect Club Network was the only firm that came to the shop, put on safety glasses, and watched a full layup before asking a single question about the market.
Founders
-
Ingrid Halvorsen
Co-founder & CEO
Composites engineer who spent a decade qualifying structural parts for aerospace programmes, and watched layup defects turn into costly NDT rework on every one of them.
-
Matteo Ferraris
Co-founder & CTO
Robotics and inspection lead behind Curelum's tow-placement autonomy, previously building automated defect-detection systems for high-rate production lines.
Our involvement
- Round led
- Seed, $3.6M
- Time to term sheet
- 11 days
- Board
- Observer seat
Since closing we have helped Curelum hire its first four robotics engineers, introduced the team to three tier-one structures suppliers through the operating network, and worked with them on the certification path that any aerospace customer will eventually require.