Feature 10
Datasets to insight
Drop a CSV or JSON file into Datasets and it becomes a live table — sortable, filterable, chartable — that can flow onward into a note or an AI chat.
Import and explore
Bring your data in as CSV or JSON and work it like a table: sort by any column, filter down to the rows that matter, and read the shape of the data before you model it.
See it, not just scan it
Chart the table to spot the pattern — distribution, trend, outlier — the moment the data lands, without exporting to another tool.
Data that travels
Send a table into a note to document it, or into a chat to ask a model about it. The dataset stays the single source; the copies are views.
In research
Turn on research mode and Datasets becomes an analysis bench: merge two separately-collected files on a participant key (with a warning when the keys don’t line up — the usual reason correlations vanish), run a data-health audit for ceiling effects, no-variance columns and small cells, build a scale with reverse-coding and Cronbach’s α, then analyze — correlations and regression with real p-values, plus reproducible Python.