/
Copy of Order Risk Scores

Copy of Order Risk Scores

Owned by Rohit Kumar
Apr 03, 2023
4 min read

What is the purpose of this article?

This article contains all the logical steps to calculate the risk score for those orders which have risk associated with the chances of failing the order and also resolving these failures at the earliest.

Figma: Order Risk Score

 

What is the approach to achieving this task?

Given below is the workflow diagram for calculating the risk score of orders and the resolution taken to handle failure.

Sequence diagram.

 

What are the steps to achieve the above task?

  1. WMS publish event on topic wms_notification

  2. A push-based subscriber (projects/eshopbox-wms-staging/subscriptions/order_created_v1_riskscore) will listen the event and push the event on Endpoint (https://order-return-dot-eshopbox-portal-dev.appspot.com/_ah/api/esb/v1/risk/order-create-listener) Event details:

    { "resource": "order", "eventType": "POST", "eventSubType": "created", "accountSlug": "tab", "accountId": null, "actor": "SYSTEM", "version": "v1", "request_data": [], "response_data": { "id": 44176830, "externalChannelID": "CH5865", "channelLabel": "Bigbasket new", "integrationType": "4", "customerOrderNumber": "Series73", "vendorOrderNumber": "Series73", "shipMethod": "STD", "orderDate": "2023-03-29 08:07:13", "expectedShipDate": null, "dispatchAfterDate": null, "promiseDeliveryDate": null, "isCOD": "0", "paymentType": "prepaid", "orderSiteID": null, "taxAmount": 0, "shipChargeAmount": 25, "subtotal": 950, "orderTotal": 975, "balanceDue": 0, "thirdPartyShipping": false, "onHold": false, "created_at": "2023-03-29 13:37:15", "updated_at": "2023-03-29 13:37:15", "shippingAddress": { "customerName": "anup", "addressLine1": "Hsidc Maruti Indl Area", "addressLine2": null, "city": "south west delhi", "state": "DELHI", "postalCode": "110022", "countryCode": "IN", "countryName": "IN", "contactPhone": "9650000000", "email": "johndoe@test.com" }, "billingAddress": { "customerName": "anup", "addressLine1": "Hsidc Maruti Indl Area", "addressLine2": null, "city": "Delhi", "state": "Delhi", "postalCode": "110761", "countryCode": "IN", "countryName": "IN", "contactPhone": "9650000000", "email": "johndoe@test.com" }, "items": [ { "lineItemSequenceNumber": 1, "orderItemID": "Series73-23924311", "itemID": "ImportTest1", "sku": "T1U973FNPT3", "productName": "Multi color jersey tee", "quantity": 1, "customerPrice": 1000, "discount": 50, "lineItemTotal": 950, "taxRate": 0, "taxAmount": 0, "giftMessage": "Happy Birthday.", "giftLabelContent": "", "cancellationReason": "", "cancellationTime": "0000-00-00 00:00:00", "original_order_item_id": null } ], "expectedDeliveryDate": null }, "previous_data": [], "resource_type": "order.create", "account_slug": "tab", "custom": [] }

  3. After listening to above event we will get account_slug based on account_slug we will query data base with the query given below:

    SELECT isRiskScoreEnabled FROM accounts LEFT JOIN account_additional_config ON accounts.id = account_additional_config.account_id WHERE accounts.account_slug = 758

  4. If the risk score enabled for given workspace, API /_ah/api/esb/v1/risk/order-create-listener will internally call API: /_ah/api/esb/v1/risk/order-verification with the request body given below:

    { "externalChannelID": "CH1299", "customerOrderNumber": "4571602911385", "externalShipmentId": "4571602911385-1299-0547", "shippingAddress": { "city": "Churcha colliery", "countryCode": "IN", "postalCode": "497339", "addressLine1": "Shreya poultry farm, Ward number 06", "addressLine2": "East nepal gate, Near balaji beauty parlour", "countryName": "INDIA", "state": "Chandigarh", "contactPhone": "9131786078", "customerName": "Anamika Dad", "email": "nomailprovided@gmail.com" }, "integrationType": "7", "items": [ { "itemID": "44372319699097", "sku": "24QGK3H1ZV5", "externalWarehouseID": "Gurgaon_FC", "quantity": 1, "outOfStock": "false" } ] }

  5. After getting request body we will get Shipping details from the request body and based on these details we calculate risk score as follows

  6. On the shipping address obtained above we will calculate the risk score, to calculate the risk score we need to verify if the address is valid or not , to verify the address we will check the Pincode, Email and Contacts respectively.

    1. Address verification: The address will be considered invalid and risk score will be HIGH. If any of the given condition is false :-

      1. Pincode should be not start with digit 0

      2. Pincode should be 6 digit long

      3. Pincode should exist in state

      4. Address length can not be lower than 60 characters including city and state. Excluding pincode.

      5. The address must include numeric values

      6. The address must not contain keywords like : “Test”, “Dummy”, “Example”, etc.

    2. Contact Details verification:- The contacts details will be consider invalid and risk Score will be considered HIGH If

      1. Contact No. is not 10 digit.

      2. Contact No. starts with country code ((after excluding last 10 digits) other than +91, 091, 91, 0)

      3. Contact No. is of pattern 1234567890 or xxxxxxxxxx (x=1,2,3,4,etc.)

      4. Customer name contains keywords like “test”,”dummy”,”example”,”123”,”Xyz”,”Abc”,etc.

    3. Customer Email verification:- Email will be considered invalid and risk score will be considered MEDIUM If

      1. Contact mail domain is incorrect. (gemail,yehoo,etc.)

      2. Contact mail contains keywords like “test”,”dummy”,”example”,”123”,”Xyz”,”Abc”,etc.

  7. Delhivery address verification API: Address Validation API, We will call Delhivery address verification API and based on its response we will assign risk score for the given order, the response body is given below, if the address-validity is:

    { "success": true, "status_code": 200, "message": "Request has been processed", "request_id": AX34687667877689, "result": { "address-validity": "Valid" } }

    1. Valid, then risk score will be LOW

    2. Ambiguous, then risk score will be HIGH

    3. Junk, then risk score will be HIGH

    4. Incomplete, then risk score will be HIGH

  8. Check for RTO score by fetching data from pincode_rto_sore table from the below query

    SELECT (total_rto/total_packed_shipmnet)*100 FROM pincode_rto_score WHERE drop_pincode ='110005'

  9. If the rto percentage is

    1. 0% then risk score will be LOW

    2. Between 0-20% risk score will be MEDIUM

    3. above 20% then risk score will be HIGH

  10. After verifying the address in next step we will check for Duplicate/Multiple orders placed by the same customers, we will assign that order risk score MEDIUM if:

    1. it is placed on the same day, with same product and same customer (same name or phone or email) and same channel

  11. After done with calculating risk score and its reasons we will save these details into database with the query given below

    UPDATE orders SET riskScore = : riskScore, riskScoreReasons = : riskScoreReasons WHERE orders.id = : id

  12. Following changes to be done in orderReturnAlgoliaModel

    1. If order is ONHOLD :

      1. For need attention, we will set riskScore, risk score reasons and onHoldReasons key into shipmentDetails in orderReturnAlgoliaModel

    2. In other cases we will set risk score details in root level in orderReturnAlgoliaModel

  13. Save the details set in orderReturnAlgoliaModel in Algolia

Code For Sequence Diagram.

 

Add label

Related content