Documentation
Technical overview
Your model runs as an encrypted artefact on customer sites, edge devices and partner clouds, and it is never decrypted there. This page collects what engineers need before a review.
Model formats
- Input
- ONNX, TensorRT
- Output
- Encrypted artefact
Gateway deployment
- Package
- Docker container
- CPU architecture
- x86, ARM
- Enclave
- Not required
- Placement
- On-prem, edge, partner cloud
Supported model classes
- CNNs
- Small to medium
- Classifiers
- Supported
- Tree models
- Supported
- Time series
- Supported
- LLMs
- Not supported
Decision cadence
- Per study, per batch, per event
- In scope
- Line-speed video
- Out of scope
Scheme parameters, operator coverage, artefact format and the key custody interface are walked through in the architecture review.
Bring one model and one target.
We tell you what it costs in latency and what it buys in protection.