Architecture and Configuration Design of the Transaction Fulfillment Order Center

The article introduces the Transaction Fulfillment Order Center, a system that records the full lifecycle of transactions—including parties, product details, amounts, billing, payment, and business information—and supports billions of orders annually across diverse business lines such as consumer finance, wealth, insurance, and payments.

Terminology : Transaction fulfillment refers to the actions taken by the seller to deliver goods or services after an order is created. The Order Center provides standardized data, full‑process queries, and downstream integration capabilities.

Overall Architecture : The system is divided into four layers—Reception, Data Processing, Data Push, and Query Service—connected via MQ messages to ensure high cohesion and low coupling. Each layer focuses on its own business logic while communicating through asynchronous messaging.

Business Integration (Configuration‑Driven) : Different business scenarios configure their own order templates and mapping rules. A unified data flow (receive → process → push) is abstracted, with conditional branches expressed via configurable expressions, enabling rapid onboarding of new business lines.

Template Configuration : Order templates define how incoming data maps to a generic order model. Data sources are routed to specific templates, and fields are extracted using JSONPath. Example code snippets:

// 解析贷款单号

Object loanId = JSONPath.extract(jsonStr, "$.jt_df_success.loanId");

// 解析还款单号

Object loanNo = JSONPath.extract(jsonStr, "$.jt_repayment.taskData.loanNo");

Expression Configuration : Conditional logic is implemented with the MVEL library, allowing lightweight rule evaluation. Example:

Object result = MVEL.executeExpression("$actExt3$=='SECOBT_JD'&&$accountType$==21", context);

Dashboard Configuration : A generic query interface and configurable dashboards let users select business lines, adjust query conditions, and customize list fields, supporting export and visualization needs.

Data Push Configuration : Processed data can be pushed to downstream systems via MQ or JEF interfaces. Push parameters are assembled from the transaction context according to configurable templates.

Future Planning : The roadmap includes enhancing configuration usability, improving system stability with circuit‑breaker and emergency response mechanisms, and expanding digital capabilities such as multi‑dimensional dashboards and alerting.

In summary, the article provides a comprehensive view of the design principles, configurable architecture, and practical implementation details of a high‑throughput backend order processing platform.