Relationships, time, movement, behaviour, statistics and language, combined in one dataset. Each area below links to a worked example you can inspect.
Many tables at once, generated in dependency order: every foreign key resolves, totals equal their line items, and dates follow each other. Start from your own DDL or a plain-English description.
See the seven-table example →Each product, store or device gets its own series with a baseline, trend, seasonality and noise. GPS tracks move along a route at believable speeds instead of jumping around the map.
See the GPS tracking example →Sessions move from step to step with realistic drop-off, built from transition probabilities. Event times stay in order and every event belongs to a real user and session.
See the funnel example →Normal, skewed, long-tail and correlated columns, derived fields calculated from other fields, and missing values, duplicates or outliers added on purpose when you need to test validation.
See the data quality example →Reviews, support tickets, notes and descriptions written for each row, so the text agrees with the rating, product and status beside it. You choose which fields use AI-generated fields; only written values are counted.
See the AI text example →232 data types, one of the largest built-in libraries of any test data generator, 222 with their own generator: VINs, IBANs, card numbers and barcodes with valid check digits, real airport and ICD-10 codes, real cities with matching postcodes. Over 200 automatic links keep each row consistent. Your own ID formats from regular expressions too. Names, emails, phones and addresses look real but belong to nobody.
See the QA example →SQL scripts load straight into PostgreSQL, MySQL / MariaDB, SQL Server and SQLite.
Every page shows one realistic dataset, checked by hand, in the tool you would open it in.
Codes, units, statuses and rules that people in the field would recognise.
Start with a free trial, or book a demo for your enterprise project.
