SIENTIAPDE-1273

Implement drift monitoring configuration and update formatters to support drift workflows. Add drift function to orchestrator utilities and enhance tests for drift functionality.

orchestrator/activities/formatters.py

@@ -19,6 +19,7 @@ with workflow.unsafe.imports_passed_through():
         minimal_retrain,
         predictions_batch,
         scouter,
+        drift,
     )
 
 topic_separator = '\n ========== \n'
@@ -34,7 +35,7 @@ class Formatters(SientiaMonitoring):
     scheduled reports.
 
     Key features:
-    - Pipeline schedule configuration formatting ("scouter", "predictions_batch", "minimal_retrain")
+    - Pipeline schedule configuration formatting ("scouter", "predictions_batch", "minimal_retrain", "drift")
     - OPC slot distribution across active ingestors
     - Notification filtering for comprehensive scheduled reports
     - Group-based report filtering with ignore list support
@@ -132,6 +133,13 @@ class Formatters(SientiaMonitoring):
                         'updated_at', now().strftime(DATETIME_FORMAT_MS_WITH_TZ)
                     ),
                 }
+            elif pipeline['workflow_type'] == 'drift':
+                schedule_config[self.laborious_namespace][pipeline['schedule_name']] = {
+                    **drift(pipeline),
+                    'updated_at': pipeline.get(
+                        'updated_at', now().strftime(DATETIME_FORMAT_MS_WITH_TZ)
+                    ),
+                }
 
         self.info('Processed schedules', metadata=metadata)
         self.debug(json.dumps(schedule_config, indent=4, sort_keys=True), metadata=metadata)

orchestrator/utils/orchestrator_functions.py

@@ -32,6 +32,25 @@ def common_config(config: dict[str, Any]):
     }
 
 
+def drift(config: dict[str, Any]):
+    """
+    Build drift configuration from pipeline config.
+    """
+    model = config['model']
+    return {
+        **common_config(config),
+        'schema': 'sientia_data',
+        'source_table_name': 'laborious_data',
+        'target_table_name': 'drift_metrics',
+        'interval': config.get('interval_minutes', 60),
+        'drift_metrics': config.get('drift_metrics', [
+            "kolmogorov_smirnov",
+            "jensen_shannon",
+            "wasserstein"
+        ])
+    }
+
+
 def minimal_retrain(config: dict[str, Any]):
     """
     Build minimal retrain configuration from pipeline config.
@@ -47,8 +66,6 @@ def minimal_retrain(config: dict[str, Any]):
     """
     return {
         **common_config(config),
-        'workflow_type': 'minimal_retrain',
-        'schedule_name': config['schedule_name'],
         'query': config['query'],
         'schema': 'sientia_data',
         'table_name': 'log_retrain',