Explore
See it work.
Every page below is a real, working surface — not a screenshot and not a recording. Open one, arrow through it, print it to PDF. They are built from the same engine and the same house style that a customer's own deliverables come out of.
These are samples, not customers. "Acme", "Acme Medical", "Cascade Bio" and "Project Northstar" are invented stand-ins. Every number, chart and result in them is illustrative — made up to exercise the components, not measured from anything real.
What is real is the output: the layout engine, the charts, the templates and the interactions are exactly what ships.
Open a tool
Two surfaces you can use right now, in the browser, with nothing to install.
DataDive
Asks one question — how deep do you want to go with your data? — and opens the workspace that matches: a look, a conversation, a clean-up, a forecast. Depths you can't use yet say so plainly instead of pretending.
The Workbench — with sample data
Drag columns into rows, columns and values; switch between a crosstab, bars and a heat map; read the profile and the quality check. It opens empty by design — the link below is you asking for a sample, and it clears in one click.
One source, two shapes
The same content rendered as a presentation and as a written report — the split that usually means writing everything twice.
Four templates, one engine
The same ten-slide deck in four designs. Nothing about the content changes — only the look, which a customer's brand can then theme further.
Worked samples
Longer decks that exercise the full component set — balance sheets, calibration plots, Gantt charts, risk registers, valuation and comps tables.
The pieces underneath
Every chart on those pages is self-contained SVG on the shared tokens — no chart library, no external fonts, printable.
Nothing here collects anything from you, and the data tools keep your rows on your device until you say otherwise.