Wens Foodstuff Group boosts R&D efficiency 10x: 12% of new code generated by Qoder CN with a 58% adoption rate

• Intelligent code completion and comment generation: Qoder CN provides real-time code suggestions and automatically generates comments for functions, classes, and modules, significantly improving coding efficiency and code readability.
• R&D acceleration: It offers real-time assistance for tasks such as interface development, business logic implementation, entity class creation, front-end code generation, and SQL script generation, reducing manual coding efforts.
• Code optimization and issue diagnosis: Qoder CN analyzes code for performance bottlenecks, identifies slow SQL, and suggests optimizations. It quickly pinpoints the root cause of interface exceptions and provides suggested fixes, dramatically reducing troubleshooting time.
• Business logic clarification and conversion: Qoder CN translates complex business rules (such as an EAS system's XML-based BOTP conversion rules) into an intuitive mapping table. It also automatically translates technical formulas into natural language descriptions, making the business logic easier to understand.
• Standardization and automation: It automatically generates standardized Java constant classes, configuration files, and other boilerplate code, ensuring consistency in naming and structure while reducing team communication overhead.

​

Key results

​

Use case 1: Significant interface performance optimization

• Query response time reduced from 3 seconds to 0.6 seconds
• Database CPU usage reduced by 45%
• Generated 5 index optimization statements, which the team validated in the production environment

SELECT
    xxx.id AS id,
    xxx AS billno,
    xxx AS xxx.alamt,
    xxx AS xxx.ear,
    xxx AS xxx.dersn,
    xxx AS wensorg,
    xxx.er AS wenscustomer,
    xxx.e AS xxx.te,
    xxx AS entryids
FROM (
    SELECT
        xxx.id AS fid,
        xxx AS fentryid,
        xxx AS fk_xxx.n,
        xxx.g AS fk_wens_org,
        xxx.er AS fk_wensxxx,
        ...
    FROM xxx
) subquery

​

Use case 2: Increased document conversion efficiency

• Mapped 132 fields, reducing the error rate from 18% to 0.The team used Tongyi Lingma's intelligent chat to clarify the field mapping from an accounts receivable document to the target document. The mapping was divided into header fields and entry fields. Header fields included org (organization), srcBillName (source bill name), description, sourceBillId (source bill ID), company (business partner), currency, bizDate (business date), id (document ID), and business partner type. Entry fields included counter account, expense type, amount payable, amount payable in local currency, exchange rate, currency, remarks, cost center, and entry ID. Some mapping rules included conditional logic, such as determining the counter account value based on an expense type code.

​

Use case 3: More efficient constant class generation

• The time to generate 800 lines of code fell from 2 hours to 5 minutes.
• Achieved 100% consistency in constant naming, reducing communication costs for cross-team collaboration.
  Copy

  /* ============================== 主实体字段 (Main Entity Fields) ============================== */
  /** xxx */
  public static final String WENS_FEMALE_STD_WT = "xxx";
  /** xxx */
  public static final String WENS_FEMALE_ACT_WT = "xxx";
  /** xxx */
  public static final String WENS_GENETICS = "xxx";
  /** xxx */
  public static final String WENS_GENDER = "xxx";
  /** 群号 (Flock Number) */
  public static final String WENS_FLOCK = xxx;
  /** 预计领用时间段 (Estimated Requisition Period) */
  public static final String WENS_TIME_PERIOD = "xxx";
  /** xxx */
  public static final String WENS_FEED_ORG = "xxx";
  /** 养户 (Farmer) */
  public static final String WENS_REARER = "xxx";
  /** 当前天龄 (Days of Age) */
  public static final String WENS_DAYS = "xxx";
  

​

Use case 4: Accelerated financial reconciliation data processing

• Processing time for 802 reconciliation data entries fell from 4 hours to 15 minutes.
• The manual calculation error rate dropped from 12% to 0.
• Troubleshooting efficiency improved significantly: interface exception investigation time dropped from 90 minutes to 10 minutes, and the tool provided multiple suggested fixes.

​

Qualitative impact

• Engineers can focus on core business logic and innovative work, such as optimizing smart farming algorithms and analyzing agricultural big data.
• The team's coding style has become more unified, improving code quality and reducing potential bugs and maintenance costs.
• This has accelerated iteration and delivery cycles for the smart agriculture and animal husbandry platform and various business systems, strongly supporting the group's digital transformation strategy.
• Qoder CN has evolved from a supplementary tool into a key productivity driver for the Wens technical team, becoming an essential AI programming assistant for every engineer.

​

Customer testimonial