Harnessing Accredo's Web Service API
What the Accredo OData v4 API actually unlocks — data integration, process automation, custom workflows — and how the Rapido SDK makes it usable without OData expertise.
2026 update. This explainer was written shortly after Rapido shipped its open-source SDK. Three years on, the SDK powers Rapido’s portal, AI processes, and every native integration we maintain. The technical primer below holds up unchanged.
Accredo’s Web Service API is the most important upgrade the platform has had in a decade. It’s a RESTful OData v4 endpoint that returns JSON, addressing the entire Accredo company file from outside the desktop application. Customers, products, invoices, orders, scripts — everything queryable, everything updateable.
It’s also genuinely difficult to use raw.
What the API unlocks
Three categories of work that used to be impossible (or required hand-rolled COM bridges):
Data integration. Sync Accredo to and from anything else — your storefront, your CRM, your data warehouse, a partner’s EDI feed. Real-time, two-way, monitored.
Process automation. Trigger workflows on Accredo events. AfterPost an invoice and reflect it elsewhere. Watch an inbox and draft orders into OE. Run reports on a schedule and email the output.
Custom functionality. Build anything Accredo doesn’t ship out of the box — a tailored sales report, a customer-specific pricing wizard, a B2B portal, an AR collections agent. The data’s all there.
Why raw OData is hard
OData v4 is versatile but verbose. Filter syntax is its own DSL. Auth has multiple modes. Field names don’t always match Accredo terminology. Errors come back HTTP-shaped instead of Accredo-shaped. Pagination, expansions, batch requests — every endpoint behaves a little differently and the spec is large enough to be its own learning project.
Most teams who try to use it directly burn a week on the basics before they get to anything useful for the business.
What the Rapido SDK does
The open-source Rapido SDK is a typed TypeScript client that wraps the OData layer with primitives developers actually want. Authentication, company-file configuration, and high-level operations against debtors, creditors, products, SQL queries, scripts, document printing, and report execution.
Result: instead of writing OData filters by hand, you write await accredo.debtors.findByName('Mainland Steel'). Instead of constructing an invoice payload from scratch, you call the typed builder.
It’s free under the Sustainable Use License for internal business, non-commercial, and personal use.
Where to start
If you want to query Accredo from a script, an internal tool, or another product:
- Read the SDK README on GitHub for installation
- Try a small read query first (e.g. fetch one debtor by code)
- Build out from there
If you’d rather not build it yourself, Rapido offers integrations into the platforms NZ businesses actually run on — Shopify, WooCommerce, HubSpot, Snowflake — plus a hosted MCP server so you can connect Claude or ChatGPT to Accredo without writing any code at all.
Talk to us about which path fits.