Zero Copy Connector for ERP Gets Smarter: The May Release Unlocks Three Ways to Build Faster

Sandeep Voona

We shipped ZCC for ERP built on SAP. That's still the core, SAP is the system of record for most enterprises. But we kept hearing the same thing from customers: "SAP is just one piece of this. We also have Workday. We have legacy systems. We have REST APIs. We want all of it in the model without building custom connectors for each."

Three capabilities in this release address exactly that.

Native REST API Support

The connector now handles any Swagger 2.0 or OpenAPI 3.0 API. Point to a public spec URL or upload a file, and the system automatically discovers endpoints.

The mechanics matter here:

Authentication isn't your problem anymore. Basic Auth, JWT, API keys, OAuth 2.0 pick one, store credentials in Vault. Token refresh happens automatically. No custom logic. No API keys in code.

Nested data works three levels deep. Real REST APIs return nested JSON objects. We map three levels in both input parameters and output fields. That handles the vast majority of what you'll encounter.

On-premises and cloud work the same way. If the API runs behind a firewall, route through a MID Server. Cloud-hosted API? Direct connection. Configuration doesn't change.

Low-code mapping. Import the spec. Use the Model Manager to visually map API response fields to your data model. No code required.

This matters because it's the gap between "we want to integrate Workday" or "we have a custom API we built" and actually doing it. Your extraction and remote tables now source from any REST endpoint. That's a meaningful expansion of what you can connect to without custom development.

An AI Agent That Discovers SAP OData Services

We shipped the OData Services Recommender. Point it at your model, and it surfaces relevant SAP OData v2 services that align with what you're building.

The problem: SAP has thousands of OData services. Most teams either miss the standard service that already solves the problem, so they build custom logic when they shouldn't or they spend days in SAP documentation searching. Both are friction.

The Recommender reads your model and suggests relevant services. It cuts discovery time and eliminates the "we didn't know this existed" moment. Your team doesn't need to be SAP experts to find the right service.

Pre-Built Models for the Employee Lifecycle

We released the Hire-to-Retire content pack. It's a set of pre-built models and process extensions that cover the employee lifecycle from recruitment through retirement.

Here's why this matters: Understanding the data model is a time sink. "How does it structure? What are the fields? How do they relate?" SAP's documentation is comprehensive, but it's not built for fast onboarding.

Content packs are read-only examples or models. Clone them, rename them, customize them.

Hire to Retire includes process extensions and higher-level abstractions that let teams use common patterns without memorizing every field.

Example: Instead of figuring out how to read active employees from SAP, query the right tables, and map the fields, you use "Read Active Employees." Clone it. Modify if you need to. Move on.

We also expanded the Enterprise Data Foundation pack with models for Vendor Bank Details, Vendor Location Details, and Vendor Contact Details. These cover the most common vendor integration scenarios we see.

Smarter Field Mapping with Visual Review

The Model Manager's mapping visualization and review interface is faster now. The system automatically suggests mappings between source and target fields, significantly reducing manual mapping time.

Better Performance on ETL Extractions

We refactored ETL data extraction from Flow Designer to a script that includes. It's a behind-the-scenes change, but it matters: extractions now run faster and more reliably.

If you're running scheduled extractions on a cadence (daily, hourly), you'll see the improvement immediately. The script-based approach is simpler, faster, and less prone to timeout issues at scale.

Content Packs Are the New Standard

We renamed "Data Products" to "Content Packs" everywhere. Same capabilities, clearer name.

You'll now see:

Enterprise Data Foundation: Core SAP data models for master data (customers, vendors, materials, etc.)
Quote to Cash: Models covering the full sales order to revenue process
Source to Settle: Models for the procurement to payment process
Hire to Retire: Models for the entire employee lifecycle (new in May)

All are templates you can learn from, clone, and customize.

What We're Building

REST APIs open up the ecosystem. The OData agent removes discovery friction. Content packs collapse time-to-model. Better field mapping removes manual work. ETL performance stops being a scaling problem.

What we're watching now is what you actually build with these capabilities. We know teams need different things. That's how the platform grows: by watching how you combine REST APIs with content packs, AI agents, process extensions, and scheduled extractions. That's where the real power is.

For more information, see Zero Copy Connector for ERP release notes on the ServiceNow documentation site.