Stop Reading About AI. Start Using It.

Realistic synthetic financial data you can throw at any AI tool. No compliance risk. No permission needed. Just $12/month.

See What's Inside a Pack

Built by a Register Accountant who got tired of waiting for IT to say yes.

syntho-bikes-2025.zip
general_ledger.xlsx319 KB
bank_statement_2025_*.mt94012 files
invoices/924 XML
income_statement.pdf21 KB
balance_sheet.pdf24 KB
management_report.pdf146 KB
trial_balance.xlsx6 KB
journal_entries.csv788 KB
948 files5.3 MB total

You know AI is changing finance.

You just can't touch it.

The compliance wall

You want to test what Claude or ChatGPT can do with a general ledger. But the moment you think about uploading real data, you already know the answer. Company policy says no. GDPR says no. Your compliance officer definitely says no.

The skills gap keeps growing

Every week there's a new AI tool, a new agent framework, a new "this will change accounting forever" post on LinkedIn. And every week, you fall further behind. Not because you're not smart enough. Because you have nothing safe to practice on.

The "Microsoft-only" cage

Your firm approved Copilot. Maybe. That's it. Meanwhile, the real breakthroughs happen in tools your IT department has never heard of. You're locked inside a walled garden while the rest of the world experiments freely.

The solution

We built the data so you can build the skills.

Complete, realistic financial datasets that look and feel like real company data. But 100% synthetic. No real companies. No real people. No compliance risk.

Realistic, not toy data

Every dataset tells a coherent story. The bank statements match the general ledger. The invoices tie to accounts payable. The trial balance actually balances. If the data isn't realistic, you're not really learning anything.

Every format you actually work with

MT940 bank files. UBL XML invoices. Excel exports. PDFs that look like they came from your client's accounting system. The same formats that land on your desk every day.

Use it anywhere

Upload it to Claude. Feed it to ChatGPT. Build an agent with LangChain. Test it in Power BI. We don't care what tool you use. You decide what to experiment with, not your IT department.

New packs, fresh scenarios

New data packs drop every month. Different industries. Different company sizes. Different accounting challenges. Your learning never goes stale.

What you get

Complete financial data. Not a demo.

Every data pack contains a full set of financial records for a fictional company. This is what separates useful test data from garbage.

General Ledger

Excel + CSV

Full chart of accounts with 12 months of journal entries. Thousands of transactions.

Bank Statements

MT940 + PDF

Standard MT940 bank files - the same format your bank uses. Transactions match the GL.

Invoices

UBL XML + PDF

Purchase and sales invoices in UBL XML (e-invoicing standard) and PDF with full VAT details.

Financial Statements

Excel + PDF

Income statement and balance sheet. Formatted PDF for presentation, structured Excel for analysis.

Management Reports

PDF

Board-level reports with KPIs, ratios, and commentary. Test AI on unstructured financial documents.

Trial Balance & Journal Entries

Excel + CSV

Raw accounting data. Every entry categorized, dated, and internally consistent.

Each pack: 900+ files, 2,000+ journal entries, 12 months of data. This is the kind of data that sits on your desk every day.

How you use it

Three ways to use your data pack

Can it actually do this?

Test AI on real accounting work

Take the general ledger, upload it to Claude, and ask it to prepare year-end adjustments. Feed the bank statements to an AI agent and see if it reconciles correctly. Find out exactly what AI can and can't do - with zero risk.

Let me show you something.

Demo AI to your team or boss

Your manager won't approve budget for AI tools based on a blog post. But show them a working demo with realistic data? That's a different conversation. Build something concrete, present it with confidence.

Finally, I can just try it.

Learn by doing, not by reading

You've read enough articles. You've watched enough webinars. The only way to actually understand what AI can do for financial work is to sit down, load real-looking data, and start prompting.

Pricing

One plan. Everything included.

POPULAR

Syntho Pro

Full access to all synthetic financial data packs

$29

$12

USD

  • All data packs included
  • Excel, PDF, CSV, MT940, UBL XML
  • New packs added monthly
  • IFRS + Dutch GAAP standards
  • Internally consistent data
  • Use with any AI tool

Cancel anytime. New packs added monthly.

FAQ

Questions you're probably asking

  • Yes. Every dataset is internally consistent - the bank statements match the general ledger, invoices tie to payables, and the trial balance balances. We built this as financial professionals, not as data scientists guessing what accounting looks like.

  • Current packs follow IFRS and Dutch GAAP (RJ). The chart of accounts, reporting formats, and VAT treatment reflect what you'd see in practice in the Netherlands and the EU. We're adding more jurisdictions based on subscriber demand.

  • Any tool. Claude, ChatGPT, Gemini, open-source models, LangChain agents, Power BI, Python scripts - whatever you want. The data is yours to use however you see fit. No restrictions, no lock-in.

  • Each pack contains a full 12-month cycle with thousands of journal entries, hundreds of bank transactions, and hundreds of invoices. Enough volume to stress-test AI tools properly - not just a handful of rows that any tool would handle easily.

  • Every month. Different industries, different company sizes, different accounting scenarios. A retail company with high transaction volumes is a very different test than a holding company with intercompany transactions. You need both.

  • At $12/month with cancel-anytime, there's not much to lose. For teams of 5+, reach out to us directly - we offer volume pricing and can customize packs for specific training scenarios. Attended one of our trainings? Use your attendee code for a discount.

The profession is changing.
With or without you.

You can keep reading about AI. Or you can start using it - today, with data that's safe, realistic, and ready to go.