Skip to content

CloudSprite overview

CloudSprite is a workspace for RF and test-and-measurement data. You store measurement files with the conditions they were taken under, compare traces in notebooks, and publish computed results back next to the originals.

How data is organized

Think in four nested containers:

  1. Organization — your company account. Billing and org-wide admins live here.
  2. Team — a lab, product line, or group. Instruments, tags, and most day-to-day work are team-scoped.
  3. Project — a campaign or DUT family (cable-qual, probe-eval). Projects hold datasets and notebooks.
  4. Dataset — one measurement event: files (Touchstone, CSV, images), parsed traces (S21, …), parameters (lot=15, temp=25), and tags.

The notebook is the canvas you actually look at. Creating a dataset also creates a notebook so there is always a place to plot that measurement. You can create more notebooks that pull traces from many datasets.

What you can do

  • Upload S-parameter and waveform files from the web app or the Python SDK.
  • Label measurements with parameters and team tags, then filter a project by those labels.
  • Overlay traces in a notebook and snapshot a view for review.
  • Run analysis in Python (including scikit-rf), then publish results as new datasets so they show up in the same project with lineage.

Per-method signatures, fields, and permissions live in the SDK reference. These how-to pages cover which object to use, in which order.

Where answers come from

Platform facts (endpoint names, required fields, who can see a project) come from these docs and the generated reference. The CloudSprite assistant follows the same rule: when a how-to is silent, it says so rather than guessing a REST path or a permission.