Skip to main content

Section 7-2 : Agentic Pattern Parallelization

How Is Agentic Pattern Parallelization Used

How Is EasyManage AI Agentic Pattern Parallelization Used For Entity level data tasks.

Entity level agentic parallel patterns are useful for parallel data tasks for one entity. e.g. Process very large table with sales transaction data, each quarter data fetched in parallel.

  • e.g. ErpSales.ErpSalesEmAgentParallel
  • Please see the log outputs below.

Prompt Parallel Workflow Example Run

  • When custom set Parallel in App:
    • Edit src\main\java\com\example\emagent\app\EmAgentSpringApp.java
    • Set String runPattern = "parallel"; //chain|parallel
2025-10-20T23:47:35.908+05:30  INFO 28280 --- [emagent_data] [           main] c.example.emagent.app.EmAgentSpringApp   : ------- Set runPattern = [parallel] ------------
2025-10-20T23:47:35.908+05:30 INFO 28280 --- [emagent_data] [ main] c.example.emagent.app.EmAgentSpringApp : ------- Run ProductEmAgentParallel ------------

Get Table Schema Response

STEP 1:
The table schema for the Product is as follows:
{
"ErpProduct": {
"type": "object",
"properties": {
"productId": {
"type": "integer",
"format": "int64"
},
"productName": {
"type": "string"
},
"productCategory": {
"type": "string"
},
"primarySupplier": {
"type": "string"
},
"productDesc": {
"type": "string"
},
"productPicture": {
"type": "array",
"items": {
"type": "string",
"format": "byte"
}
}
}
}
}

Get Table Data Response

STEP 2:
## Table Data

The data you gave has these fields, but actually includes a **primarySupplier** which is not in the schema (this might be metadata or from a related table), and lacks values for Price, StockQuantity, CreatedAt, UpdatedAt.

Sample data:

ProductID | ProductName | Category | PrimarySupplier |
-----------|-------------------|-----------------|-----------------|
101 | Smartphone Model X| Electronics | Supplier A |
102 | Vacuum Cleaner Pro| Home Appliances | Supplier B |
103 | Learning JSON | Books | Supplier C |

Analyze For Insights Response

STEP 3:
  • Step 1 and Step 2 are excuted in Parallel.
  • Upon both completion, Step 3 is run.