What you can do in Data Aqmen
One workspace where data, the SQL that shapes it, the charts that show it, and the conclusions drawn from it all stay connected — and every piece keeps its receipts.
01 — Provenance
Datasets that remember where they came from
Drop a CSV, JSON, or Parquet file, check the parsing preview, and land it as a versioned table. Later files append or replace — history is kept either way.
Every dataset carries an editable schema — types, units, descriptions — and a sources list: citations for where the data came from and the judgement applied preparing it. Your numbers stay defensible.
| company | segment | raised |
|---|---|---|
| Personio | Core HR | €500M |
| Deel | Payroll | €425M |
| Factorial | Core HR | €120M |
2 sources
02 — Lineage
Transform with SQL, trace every step
Write a SELECT, name the output, run it — that's a transformation: a clean table materialized from your raw data, with the result, the schema, and the SQL recipe on one page.
When an input changes, the output is marked stale; one click re-runs it. The lineage graph maps sources to datasets to charts, and can run the whole pipeline in dependency order.
fresh
03 — Queries
Charts that are never out of date
Explore in the SQL editor with the schema browser at your side — ⌘⏎ runs, results export to CSV, and the editor is read-only by design, so you can never damage the data.
Save the keepers as queries with a chart. They re-run when opened, so what you see is always the current data — never a stale export.
Funding by segment
re-run on open · current as of today
core
pay
ats
l&d
wfm
04 — Views
Dashboards you describe, not build
A view is a small interactive page over your live data — several charts, filters, a layout — built by an agent from a plain-language description. Say the screen you want; refine it in the same conversation.
Views run on the workspace itself: every number is a query against the current datasets, so they stay live and inspectable. And they stay focused — one view per question, not a wall of widgets.
"A view of funding by segment with a quarterly trend and top deals"
Funding overview
view · live data€2.4b
214
€8.2m
every number is a query against current datasets
Agent-first
Or skip the clicking — ask.
Load the latest EU HR-tech funding rounds — cite every source.
Which segment grew fastest since 2021? Chart it.
Re-check my insights against the new data.
Your AI agent connects straight to your workspace. Everything it does lands as real artifacts — datasets, queries, insights — cited, filed, and auditable.
05 — Insights
Conclusions with receipts
Insights are short claims pinned to the query or dataset that demonstrates them. Validate the ones that hold, reject the ones that don't, reopen when circumstances change.
When the data or SQL underneath a claim moves, it's flagged so you know to re-check — analysis stops silently rotting.
Core HR captures 62% of tracked funding
evidence: funding_by_segment
06 — Collections
One question, one folder
Collections group the datasets, queries, and views that belong to one project or question, nested up to three levels, always in reach in the sidebar.
Wherever you save something, a picker files it in the right place — and ⌘K search finds anything, filed or not.
07 — Activity
Nothing happens off the record
The Activity feed is an append-only record of who did what: imports, transformation runs, query saves, insight edits. Searchable, and every entry clicks through to the thing it touched.
Agents work in the same workspace, so their actions land in the same feed — nothing that happens while you're away is invisible.
elena imported dataset companies · 20 rows
2h agomarcos ran transformation funding_by_segment
1h agomarcos recorded insight “Core HR captures 62%…”
1h ago