AI Associate Data for AI — All Questions With Answers

{
  "policy": {
    "resource": "customer_data",
    "action": "read",
    "conditions": [
      {"field": "region", "operator": "eq", "value": "EU"}
    ],
    "masking": {
      "fields": ["email", "phone"],
      "method": "partial"
    }
  }
}

2023-10-01 12:00:01 ERROR [DataPipeline] com.salesforce.datalake.pipeline.TransformException: Field 'account_id' not found in schema. Expected String, got null.
2023-10-01 12:00:02 INFO [DataPipeline] Retrying task 3/3 after 5000ms.

{
  "maskingPolicy": {
    "name": "MaskSSN",
    "fields": ["SSN", "CreditCard"],
    "maskingType": "partial",
    "character": "*",
    "showLastFour": true
  }
}

2025-03-01 14:32:15 ERROR [DataTransformRunner] Transform failed: java.lang.ArithmeticException: / by zero
at com.salesforce.dc.datatransform.FormulaEvaluator.processField(FormulaEvaluator.java:256)

SELECT Campaign__c, COUNT(*) as Records, COUNT(Response__c) as NonNullResponse
FROM UnifiedCustomer
GROUP BY Campaign__c

Results:
Campaign__c | Records | NonNullResponse
Summer Sale  | 1000    | 750
Spring Promo | 800     | 800
Fall Clearance | 600    | 0

{
  "fieldMapping": [
    {"sourceField": "Id", "targetType": "Text"},
    {"sourceField": "AccountName", "targetType": "Text"},
    {"sourceField": "CloseDate", "targetType": "Date"},
    {"sourceField": "Amount", "targetType": "Number"},
    {"sourceField": "LeadSource", "targetType": "Category"}
  ]
}

Date,Clicks,Conversions
2023-01-01,100,10
01/02/2023,150,15
2023-03-01,200,

Total records: 10000
Missing values: 500
Duplicates: 200
Outliers in Amount: 50

{
  "dataAccessPolicy": {
    "dataset": "customer_transactions",
    "allowedUsers": ["data_scientist", "ml_engineer"],
    "deniedUsers": ["intern"],
    "fields": ["transaction_id", "amount", "date"],
    "conditions": {
      "amount": {"gt": 0, "lt": 10000}
    }
  }
}

2024-05-10 14:32:15 ERROR PipelineRunner - Stage: DataTransformation
2024-05-10 14:32:15 ERROR PipelineRunner - Exception: Column 'age' not found in schema
Schema: name string, income float, region string

transform:
  - type: one-hot
    columns: [color]
  - type: standard-scaler
    columns: [price, weight]

{
  "model": "EinsteinDiscovery_Churn_v2",
  "status": "TRAINING_FAILED",
  "errorCode": "INSUFFICIENT_POSITIVE_EXAMPLES",
  "fieldCount": 8,
  "recordCount": 3200,
  "positiveExamples": 180
}

SELECT AccountId, SUM(Amount) TotalAmount
FROM Opportunity
WHERE CloseDate > LAST_N_DAYS:365
GROUP BY AccountId
HAVING TotalAmount > 100000

{
  "dataflow": "EinsteinDataPrep_Sales",
  "nodes": [
    {
      "id": 1,
      "type": "source",
      "object": "Opportunity"
    },
    {
      "id": 2,
      "type": "transform",
      "operation": "filter",
      "condition": "StageName = 'Closed Won'"
    },
    {
      "id": 3,
      "type": "output",
      "target": "EinsteinDiscovery_Stage"
    }
  ]
}

Refer to the exhibit.

Error log:
2023-11-01 10:23:45 ERROR: Failed to load data from source 'salesforce_opportunities'. Reason: Field 'Amount' has null values in 30% of records.
2023-11-01 10:23:46 WARN: Data quality check: 'Stage' field contains 15 distinct values, expected 8.

Refer to the exhibit.

JSON policy:
{
  "dataAccess": {
    "allowedSources": ["Salesforce", "AWS S3"],
    "dataFields": ["Account.Name", "Contact.Email"],
    "dataMasking": {"SSN": "XXXXX"},
    "retentionDays": 90
  }
}

{
  "dataset": "OpportunityPrediction",
  "fields": [
    {"name": "Amount", "type": "currency", "missing": "fill_with_median"},
    {"name": "Stage", "type": "picklist", "missing": "exclude"},
    {"name": "CreatedDate", "type": "date", "missing": "use_default"},
    {"name": "Description", "type": "textarea", "missing": "ignore"}
  ],
  "model": "EinsteinDiscovery",
  "target": "Won"
}

Pipeline: MyPipeline
Status: FAILED
Error: RecordsProcessed: 5000, Errors: 120
LastError: FIELD_INTEGRITY_EXCEPTION: CustomLeadField__c: value not found in picklist values

{
  "modelConfig": {
    "modelType": "classification",
    "targetField": "Churn__c",
    "featureFields": ["Age__c", "Tenure__c", "Usage__c"],
    "dataSource": "DataCloudObject",
    "trainingWindow": "Last_90_Days",
    "predictionWindow": "Next_30_Days",
    "splitRatio": 0.8,
    "regularization": "L2"
  }
}