Connecting EZcount (איזיקאונט) to AI and Automations via Its API

If you run your invoicing on EZcount (איזיקאונט), you're in a better position than most Israeli businesses when it comes to automation. Unlike some legacy accounting tools that hide behind file exports, EZcount ships a documented REST API and official SDKs for PHP, Java, .NET, ASP, Python, Ruby and Node.js. That means an AI workflow doesn't have to scrape screens or import CSVs — it can talk to your account directly and reliably.

How the EZcount API actually works

EZcount exposes a straightforward REST API. You authenticate every request with an API key plus a developer email you generate from your account settings, and you POST JSON to endpoints. The workhorse is createDoc (and createDocV2), which issues a tax invoice, receipt, invoice-receipt, quote or credit note — and can also clear a credit card in the same call, so you can charge and document in one step. There's a demo base URL for testing before you point at production.

• Auth: an api_key plus your developer/API email on every request — no OAuth dance, which keeps server-to-server jobs simple.
• Create: createDoc / createDocV2 to issue documents and optionally charge a card or save a token (no PCI burden on your side).
• Read: search and fetch documents to pull what was already issued — the foundation for reconciliation and reporting.
• Debug: a debug_key flag plus getLastAPIRequestData lets you see exactly what the server received — a small detail that saves real hours.

One honest caveat: EZcount's public API is request-driven (you call it; it answers). It does not advertise a rich webhook event stream the way Stripe does. So an AI integration usually polls on a schedule — "every 15 minutes, fetch documents created since the last run" — rather than receiving instant push events. For invoicing workloads that's almost always fine; it just shapes how you design the automation.

What you can actually build

1. Auto-issue invoices from any source: a Shopify/WooCommerce order, a form, or a CRM deal closing fires createDoc and emails the customer a proper Israeli tax invoice — no manual typing.
2. AI reconciliation: an agent pulls the day's EZcount documents and your bank/payment-processor feed, matches them, and flags only the discrepancies for you to review.
3. WhatsApp billing replies: a customer asks "שלחתם לי חשבונית?" and an AI agent looks up their real EZcount document and replies with the link — using actual data, not a guess.
4. Smart alerts: AI watches issued documents and pings you when something looks off — a duplicate, an unusually large amount, or a client who suddenly stopped paying.
5. Month-end summaries: an LLM reads the period's documents and drafts a plain-language revenue summary for you and your accountant.

EZcount gives you a clean API. The value isn't issuing one invoice from code — it's the reliable layer that turns your real billing data into automations you trust.

No-code vs. custom code

If your flow is simple and low-volume — "new order, make invoice, email it" — a no-code tool like Make can call the EZcount API directly and you may never need a developer. The line gets crossed when you need real logic (partial payments, refunds, multi-currency, VAT edge cases), higher volume, retries that don't double-issue documents, or AI reasoning over the data. That's where a thin custom service beats a brittle chain of automation steps.

I build exactly this kind of integration — connecting EZcount to AI and to the rest of your stack, end to end, with the boring reliability work (idempotency, retries, logging) done properly so you never double-charge a customer or lose a document. If you're looking for a developer to connect EZcount to AI and automations via its API, that's the work I do. The contact form on this page reaches me directly — tell me what you're issuing today and what you wish happened automatically.