JD Supply Chain Low-Code Rule Engine Platform: Architecture, Challenges, and Solutions

The article introduces a low‑code rule engine platform created by JD Supply Chain's R&D department to address the limitations of open‑source rule engines in high‑traffic internet scenarios, such as insufficient concurrency handling and difficulty for non‑technical users.

It explains how the platform, named HaiNa, enables business and development roles to collaboratively design, deploy, and manage business rules through a visual designer, while the backend automatically generates rule description files that are executed via an SDK in the target application.

Key challenges identified include performance (handling billions of orders per day with nanosecond‑level rule execution) and value (providing a domain‑specific language that is accessible to product and business users). The solution separates technical and business concerns, offering lightweight integration (SDK + annotation) and high‑performance, decentralized execution.

Core features highlighted are:

Lightweight integration requiring only the SDK and annotations.

High performance validated during JD's 11.11 promotion, supporting over a hundred million orders per day.

Clear separation of development and business roles, allowing non‑technical users to compose rules visually.

Visual, drag‑and‑drop rule composition resembling flowchart design.

Technical asset accumulation through reusable rule libraries.

Practical examples demonstrate rule registration, business model registration, technical rule registration, and rule execution flow, emphasizing the platform's ability to reduce delivery cycles from ten days to under a day.

In conclusion, the platform empowers both developers and business stakeholders to achieve rapid, self‑service rule deployment, dramatically improving delivery efficiency and scalability across JD's core fulfillment systems.