Data Integration Glossary Part 1: iPaaS, ETL, API, Webhook and 6 More Core Concepts

Why a Glossary

In integration projects the business says "import the orders", the architect says "build an incremental sync pipeline", and the developer says "call their open API" — three vocabularies for the same job. Part 1 of this glossary covers the core concepts of integration forms and technical infrastructure.

The Terms

iPaaS (Integration Platform as a Service) — a cloud platform that replaces hand-coded point-to-point interfaces with pre-built connectors, visual flow orchestration and unified operations. Every new system pair becomes a configured data flow instead of a new codebase. Qeasy is a domestic iPaaS focused on ERP, e-commerce and supply-chain scenarios.

ETL (Extract, Transform, Load) — the classic paradigm: cleanse and transform data in transit so the target receives clean records. ELT (Extract, Load, Transform) — land raw data first and transform inside the warehouse, leveraging warehouse compute. Business-system synchronization (orders, inventory) is almost always ETL thinking; ELT suits analytics.

API (Application Programming Interface) — the programmatic contract a system exposes; the fundamental channel of data integration. Webhook — an event-triggered reverse HTTP callback: "call me when something changes." The two are complementary in production: webhooks deliver real-time events (order changes), while API polling provides backstop and initialization (scheduled full reconciliation). Webhooks alone drift due to message loss; polling alone fails on both latency and call volume.

SDK (Software Development Kit) — the platform-vendor's official client library that hides signing, serialization and retry details.

API Orchestration — real business rarely completes in one call: "create a sales order" may require looking up the customer, then the product, then creating the order, then writing back the order number. Orchestration defines such call chains with branches and error handling as a reusable flow — a core iPaaS capability.

Field Mapping — the rules translating source fields to target fields, including value transformations such as mapping platform status codes to ERP status enums. Data Cleansing — fixing or rejecting missing, duplicated or malformed data. Together they determine the trustworthiness of integrated data and consume the bulk of implementation effort.

Middleware / Connector — a connector is a pre-built integration component for a specific system, bundling authentication, endpoints and data models; middleware is the broader infrastructure layer such components run on.