Developer Tools
Platform SDKs
The SDK layer exposes the same lifecycle model the rest of the platform uses. Its purpose is not only convenience. It is to let teams bring datasets, workflows, releases, fleet actions, and operational evidence into notebooks, CI jobs, internal services, and robot-side code without breaking lineage or governance.
SDK Surfaces
Python SDK
The main surface for research, data, training, evaluation, and automation workflows.
- dataset and benchmark access
- curation and finalization actions
- training and experiment launch
- model and release inspection
- operational query and automation
Robot Agent SDK
The edge-side surface for data upload, telemetry, artifact delivery, and controlled update behavior.
- session registration and upload
- telemetry and status streaming
- rollout and update coordination
- local resilience for field conditions
Shared Design Principles
Same entities as the platform
SDKs operate on the same dataset IDs, run IDs, model IDs, artifact IDs, fleet IDs, and workflow IDs the UI and API use.
Same governance model
Programmatic actions still respect access control, approvals, and auditability.
Same adoption story
Teams can keep notebooks, scripts, internal tools, and robot software while still moving lifecycle state into rFabric.
Why Teams Care
Developer trust
Teams adopt platforms more deeply when they can script them and embed them into existing engineering workflows.
Faster automation
CI jobs, research notebooks, and robot agents can act directly on first-class platform objects.
Cleaner lineage
Programmatic actions no longer require exporting data out of the system and losing lifecycle continuity.