Protect your model. Wherever it runs.
Your classifier ships inside an OEM machine or onto a customer's line. It stays encrypted there: no plaintext in memory, nothing to copy off the industrial PC.
Where the model walks away today
01The OEM deal abroad
A partner in Asia or the US asks for source escrow or perpetual runtime rights before signing.
02The customer who wants to self-host
The end customer wants your model on their own servers, permanently, after the contract ends.
03The cloned machine
Mechanics can be copied. The trained classifier is the one asset a copyist cannot build, unless they lift it from the box.
Where it runs
- Industrial PCsGPU and Jetson-class
- Serversx86 on the customer line
- PackageDocker container
- Jetson-class latencyMeasured on request
Fit with your specific vision pipeline is mapped in the architecture review.
Keep your dongle
Licensing tools
Entitlement: who may run the model, where and for how long.
Wodan
Confidentiality during inference.
They work side by side.
Pilot
€20K to €50K
Indicative range, depending on model and target. The fixed fee is confirmed in writing after the architecture review.
- Your model converted and running encrypted on your target hardware
- Benchmark report on your hardware: latency, accuracy delta, memory
- Deployment recipe
Bring one model and one target.
We tell you what it costs in latency and what it buys in protection.