Pilot
One model. One target. A number you can show your board.
A fixed-fee pilot: we convert one of your models, run it encrypted on your target hardware and give you the measurements and a decision memo for your approver.
Steps
- 01
Architecture review
- 02
Scope and fee
- 03
Convert the model
- 04
Run encrypted on your target
- 05
Benchmark report and decision memo
One model, one target
What you get
- Your model converted and running encrypted on your target hardware
- A benchmark report: latency, accuracy delta and memory, on your hardware
- A deployment recipe
- Threat-model fit for your deployment
- A decision memo for the approver
What you provide
- The model (ONNX or TensorRT)
- Representative inputs
- A target device or VM
- One engineer for two hours a week
Bring one model and one deployment.
We map which layers fit your hardware and risks, and what a paid pilot would confirm.
Our engineers review your model and deployment context with you.