Enrich spreadsheets

Input formats

enrich reads CSV, TSV, JSON, JSONL, and XLSX. The input format is detected from the file extension.

Enrich a file

Point enrich at a file, choose the services, and write the result with -o:

demografix enrich people.csv -o out.csv --gender --age --nationality --name-col full_name

Each enabled service appends its own columns. --gender adds gender, gender_count, and gender_probability. --age adds age and age_count. --nationality adds ranked country candidates and nationality_count.

Choose the name column

enrich needs to know which column holds the name. Name it with --name-col, using a header name or a 1-based column index:

demografix enrich people.csv -o out.csv --gender --name-col 2

When first and last names are in separate columns, use split mode:

demografix enrich people.csv -o out.csv --gender \
  --first-name-col first --last-name-col last

Scope by country

Gender and age accept a country. Apply one country to every row with --country, or read it per row from a column with --country-col:

demografix enrich people.csv -o out.csv --gender --age --name-col full_name --country-col country

Nationality candidates

Nationality returns several country candidates per name, ordered by descending probability. Set how many are appended with --top-n (1 to 5, default 3):

demografix enrich people.csv -o out.csv --nationality --name-col last_name --top-n 5

This appends country_1 through country_5, a matching country_N_probability for each, and nationality_count.

Avoid column collisions

A fresh run refuses to overwrite an existing column. If a prediction column name already exists in the input, add a prefix with --prefix:

demografix enrich people.csv -o out.csv --gender --name-col full_name --prefix pred_

Output

For enrich, -o is an output file path, not a format. The output format follows the file extension — .csv, .tsv, .json, .jsonl, or .xlsx.

Preview the cost

--dry-run validates the configuration and prints a plan — input rows, appended columns, and the number of names that would be billed — without calling the API:

demografix enrich people.csv -o out.csv --gender --age --name-col full_name --dry-run

plan
  input        csv, 1200 rows
  output       out.csv
  services     gender, age
  name column  full_name
  country      none (global)
  new columns  gender, gender_count, gender_probability, age, age_count
  cost         2400 names — no API calls made

Each enabled service bills one name per non-empty row.

Resume after an interruption

A failed request does not abort the run. Sibling rows still process, and the partial output is written. If the quota runs out mid-run, enrich reports how many rows it wrote.

Re-run with --resume to fill only the rows whose prediction columns are still empty:

demografix enrich out.csv -o out.csv --gender --name-col full_name --resume

Summarize a signup file

Enrich a file of signups, then aggregate the result:

demografix enrich signups.csv -o signups_enriched.csv \
  --gender --age --nationality --name-col full_name

The output adds gender, age, and nationality columns to every row. Pivot it in a spreadsheet or notebook to report the demographic mix of the cohort — the gender split, the age distribution, and the leading nationalities.

To predict a few names without a file, or to pipe a list of names, see Basic usage.