What’s in a dataset
A dataset doesn’t need a target column. Predictive components (like the predictor) take the target column as a typed input when you configure an experiment. See Experiments.
What you do with a dataset
Three primary actions, in increasing order of commitment:- Chat — open Summand and ask questions. The dataset’s curated Parquet, schema, column stats, and context are all available to the assistant. It can run one-shot SQL, generate charts, and propose views or experiments.
- Create a view over it — save a SQL transformation, optionally joining other datasets, queryable by chat and usable as an experiment source.
- Set up an experiment — schedule one or more components to run on the dataset (or a view of it) on a cron. Outputs are versioned.
The dataset detail page
Datasets have a configuration page with these tabs. They’re all about tuning behavior, not browsing analysis output (analysis output lives in chat and on the Components tab):
Most users never visit anything beyond Overview.
Schema and types
Summand infers column types from the data on ingest. The Schema tab shows the inferred types and lets editors add a description to each column.Versioning
Each component run produces a new computed-context version keyed by(datasetId, version). The latest version is what the UI reads; older versions remain queryable for audit and comparison.
A new version is created when components run:
- You start a run from the dataset page or from chat
- An experiment runs (manual or scheduled)
- A retrain policy fires on its schedule, on drift, or when enough new rows arrive
Limits
The active-dataset cap applies per account, not per connector:- Free: 3 active datasets total
- Pro / Education: unlimited
- Enterprise: unlimited