Import & sources

Every number in Aqmen should be traceable to a document and readable by a teammate or an agent. Two things make that true: sources, the documents the work stands on, and datasets, the clean tables taken from them, each citing its sources and saying what its columns mean.

Data arrives clean: Aqmen is not a cleaning tool. Prepare the file so it is ready to use — your agent can do this for you — then import it.

Sources

A source is a document the work stands on: a file (a report, a filing, a PDF, a spreadsheet), a public link, or a reference such as an expert interview, a meeting or an estimate. On Data, choose New → Source… and pick File, Link or Reference.

A dataset cites a source, saying where in it the numbers were read and what was done while reading them.

Importing a file

On Data, choose New → Data… → Import a file. The import takes three steps:

1. File and destination: pick a CSV, TSV, JSON or Parquet file (gzipped works too), and choose to create a new dataset, or append to or replace an existing one. Upload & preview.
2. How to read the file: for a CSV, set the delimiter, whether the first row is a header, and rows to skip, then Apply & re-read until it looks right.
3. Columns: choose the columns to keep, rename them, and check their types against a preview of the first rows. Under Where is this from?, Cite a source. Then Import — nothing lands until you do.

To add to a dataset later, open it and use ⋯ → Append data… (rows are matched by column name) or Replace data….

The dataset page

• It opens on the data, with numbers formatted from each column's unit.
• Edit shows how it is described: each column's meaning, unit, scale and description — what makes a column called amt legible as "funding amount, EUR millions" — and the sources it cites. Changes save as you make them; Done goes back to the data.
• Details shows what reads this dataset — transformations, charts, spreadsheets, views — and its notes, which agents read before using it.

On Data, a dataset that cites no source says so: No source cited.

Why citations matter

A table of numbers with no origin cannot be defended in a meeting, and neither can an agent reasoning from it. Datasets your agent loads arrive with their sources cited; the ones you import deserve the same care.

Faster with an agent — ask your agent to "research <topic>, clean the results, and load them as a dataset with sources." It gathers the data, stores the documents as sources, loads the clean table, describes its columns and cites every origin.