Skip to content

First launch

Scelo opens on the pipeline: soft data, tools, hard data

The pipeline, straight away

Every launch plays a short intro, the S₀.₂ mark drawing itself in, and opens on the Scelo pipeline: the macro view of the three stages, Soft → Tools → Hard (see The pipeline). get started, top left, takes you to the welcome screen. The bar above the pipeline carries:

  • workspace — the files, editor and terminal (see The workspace). With no workspace open yet, it takes you to the welcome screen to pick one.
  • open .sce / save .sce — the whole session as a project file.
  • export · whole pipeline.
  • + start project — name the session so its chats persist.
  • reset session — wipe the dataset, model picks, runs and activity log. It asks first.

A click or a key skips the intro; with reduced motion on, the mark simply appears.

The runtime check

The runtime check reports on the bundled stack:

  • Python — the portable interpreter version and that it can import the IA package set (numpy, pandas, lifelib, chainladder, climada, …).
  • R — the portable R version and that the actuarial libraries (ChainLadder, chainladder, lifecontingencies, forecast, …) load.
  • Status per component — green when ready; a clear message if something didn't stage.

Open it any time with Navigate: Runtime Check in the command palette, or at the /runtime-check route. If a component shows an error, see Troubleshooting.

The welcome screen

The welcome screen — open a folder or scaffold a sample workspace

The welcome screen is where you choose where to work. get started, top left of the pipeline, always opens it; workspace takes you there while no workspace is open, and the workspace header links back to it:

  • Open Folder… — point Scelo at any directory; it becomes the workspace root.
  • Switch Workspace… — jump between workspaces Scelo knows about.
  • Configure AI Provider — Ollama is the local default (no key, no spend); switch to a hosted provider here. See AI providers.
  • Download a dataset — grab a starter dataset (IBTrACS cyclones, WHO life tables, NFIP claims, ChEMBL, …).
  • Sample workspaces — one-click scaffolds you can copy to disk to learn the flow:
    • Life pricing starter — mortality table + deterministic premium walk in Python and R.
    • Climate risk starter — IBTrACS + Climada loss curve.
    • Scelo brain starter — a minimal soft → tools → hard pipeline.
    • Reserving starter (Mack chain-ladder) — RAA triangle through R ChainLadder + chainladder.py, cross-checked.

Where to next