Demo Data
Demo data that doesn’t look like demo data.
Paste one realistic row. Get hundreds of distinct, plausible, PII-free rows in CSV, JSON, and Postgres SQL — for screenshots, sales calls, demo days, and staging databases that have to look real.
[
{
"id": 2,
"full_name": "Marcus Kowalski",
"company": "Velocity Analytics",
"plan": "Enterprise",
"mrr_usd": 299,
"signed_up_at": "2025-09-17T09:45:33Z"
},
{
"id": 3,
"full_name": "Priya Nakamura",
"company": "Cascade Design Studio",
"plan": "Starter",
"mrr_usd": 29,
"signed_up_at": "2025-12-02T16:11:08Z"
},
…498 more rows
]How it works
Paste a sample row
One realistic JSON object — keys, types, and any enum-like values are inferred from that one example.
Pick a row count
Up to 500 rows in the starter pack. Live preview shows 5 free rows before you pay.
Pay $19
Stripe Checkout — one-shot. No subscription, no account.
Download CSV / JSON / SQL
We generate in batches, validate every row against your schema, and email you a download link within minutes.
What you get
Realistic, varied rows
Names look like names, not Smith and Doe. Companies look plausible. Dates fall in sensible windows. Statuses respect the enums implied by your sample. The generator is LLM-driven, so a pack of 500 rows is genuinely 500 different rows — not the same dozen names rotated.
Schema from one sample row
Paste a single realistic row of JSON. We infer keys, types, and enum-like values from that one example and produce N rows in the same shape. No field-by-field type configuration, no schema editor — the example is the schema.
Coherent, demo-grade output
Amounts cluster in plausible ranges, dates flow forward sensibly, and status fields stay inside the obvious set. The output is intended to make a screenshot or sales demo look real, not to fail a glance test.
Privacy-safe by construction
Every row is synthetic. Your sample row is used only to infer the shape and is wiped after the generation run. Demos, App Store screenshots, blog posts, and staging environments all stay clear of real PII.
Built for demos that have to look real
- Y Combinator / Techstars demo prep
- App Store and Product Hunt screenshots
- Sales-call dashboards with “your biggest customer” energy
- Staging databases that mimic production volume without the PII risk
- Launch-week blog posts that need a screenshot of a full product
Three formats, one pack
CSV
Header row plus rows, ready for spreadsheets, BI tools, or bulk imports.
JSON
An array of objects matching your sample row's shape, pretty-printed.
Postgres SQL
INSERT INTO ... VALUES (...) statements with column types inferred from your sample.
Generate a starter pack
The full generator opens shortly. Until then, drop your email and we will send you a free 50-row sample of any of our starter schemas (customers, orders, deals, products, episodes), so you can see the output before you decide.
Reach us directly at demo-data@forage.bot.
Questions
- What do I send you, exactly?
- One realistic JSON row that looks like the data you want. We infer the keys, the value types, and any enum-like values from that one row, then generate the number of new rows you ask for in the same shape.
- Is this real customer data?
- No. Every row is synthetic. Your sample row is used only to infer the shape of what you want; it is passed through the generator once and then discarded.
- What's the row cap?
- The starter pack delivers up to 500 rows for one schema. If you need more, contact us before purchase.
- How is this different from Faker / Mockaroo?
- Faker and Mockaroo are rule-based generators where you configure each field by type. Demo Data infers the shape from one example row and uses an LLM to produce varied, realistic-feeling values, so a pack of 500 rows looks like 500 distinct rows rather than a hundred Smiths.
- What if the output doesn't match my schema?
- If the assembled output does not match the schema you supplied, or fewer rows than you ordered are delivered, we issue a full refund within 7 days. See Terms.
- Can I use this for anything other than demos?
- Use it anywhere synthetic, schema-shaped data is genuinely helpful — staging environments, screenshots, blog post examples, demos. Do not use it to impersonate a real person, fabricate evidence, or seed reviews. See Terms.