Not random noise — correlated, business-realistic datasets. Pick a domain, set a seed for reproducible output, preview it, and export CSV or Excel. Everything runs in your browser; no data ever leaves your machine.
Invoice line items driven by customer segments (A–D), relationship momentum, large-buy spikes, and category markups. The original simulator that started this project.
An MRR movement ledger — new, expansion, contraction, churn — with plan tiers, seat counts, and plan-dependent churn. Built for NRR & cohort modeling.
Order lines with RFM-style segments (Champion → At-Risk), weekly + holiday seasonality, promos, and realistic return rates.
Basket-level transactions with real market-basket affinities (chips + salsa + soda), store IDs, hour-of-day patterns, and a payment mix.
Most mock-data tools fill columns with independent random values, so the rows have no internal logic. These generators simulate the process behind the data: customer segments, buying behavior, seasonality, churn, basket affinities and pricing all interact, so the output behaves like a real export. That makes it genuinely useful for BI demos, dashboard tutorials, data-science take-homes, load testing, and teaching.
Browse shared dataset recipes — each opens a generator pre-loaded with a fixed seed and settings, so you get the exact same dataset as everyone else. No data is uploaded; a recipe is just a seed plus a config. Generate your own on any tool page, then Copy shareable link or Export recipe (.json) to share it back.
No. All generation happens locally in your browser via JavaScript. Nothing is sent to a server, which is why you can safely generate large files and use it on sensitive networks.
You share a recipe — the seed and settings — as a link or a small .json file. Because generation is deterministic, anyone who opens it reproduces the identical dataset on their own machine. Privacy intact, datasets still shareable.
The seed makes output reproducible. Enter the same seed, row count, and options and you get a byte-identical dataset every time — ideal for tutorials, bug reports, interview questions, and version-controlled demos. Leave it blank for fresh random data.
Yes — it's entirely synthetic and contains no real people or companies. Use it for demos, products, courses, or testing.