PLANNED PORTAL · PRODUCT 0%

Discovery experiment workspace

This is the specification for the future product workspace. No uploads, model runs or physical experiments are executed here.

Inputs

A documented dataset, one prediction target, split strategy and model configuration.

Intended workflow

  1. Select a versioned benchmark.
  2. Inspect data quality and train-test separation.
  3. Run a baseline and compare a candidate model.
  4. Export rankings with uncertainty and experiment metadata.

Outputs

Benchmark metrics, model cards and reproducible experiment bundles.

Opening criteria

Open after a baseline experiment can be independently reproduced within an agreed tolerance.

Current availability

The strategy and public changelog are available. The interactive product has not been built yet.

Read the build strategy