feat(e2e-tests): added inference of pipeline

tests/integration/godog/features/model/custom_model_deployment.feature

@@ -19,7 +19,7 @@ Feature: Explicit Model deployment
       storageUri: gs://seldon-models/scv2/samples/mlserver_1.3.5/iris-sklearn
     """
     When the model "alpha-1" should eventually become Ready with timeout "20s"
-    Then send HTTP inference request with timeout "20s" to model "alpha-1" with payload:
+    Then I send HTTP inference request with timeout "20s" to model "alpha-1" with payload:
     """
     {
         "inputs": [
@@ -52,7 +52,7 @@ Feature: Explicit Model deployment
     }
   ] }
     """
-    Then send gRPC inference request with timeout "20s" to model "alpha-1" with payload:
+    Then I send gRPC inference request with timeout "20s" to model "alpha-1" with payload:
     """
     {
         "inputs": [
@@ -85,7 +85,7 @@ Feature: Explicit Model deployment
   ] }
     """
     Then delete the model "alpha-1" with timeout "10s"
-    Then send HTTP inference request with timeout "20s" to model "alpha-1" with payload:
+    Then I send HTTP inference request with timeout "20s" to model "alpha-1" with payload:
     """
     {
         "inputs": [
@@ -99,7 +99,7 @@ Feature: Explicit Model deployment
     }
     """
     And expect http response status code "404"
-    Then send gRPC inference request with timeout "20s" to model "alpha-1" with payload:
+    Then I send gRPC inference request with timeout "20s" to model "alpha-1" with payload:
     """
     {
         "inputs": [

tests/integration/godog/features/model/over_commit.feature

@@ -51,7 +51,7 @@ Feature: Explicit Model deployment
       storageUri: gs://seldon-models/scv2/samples/mlserver_1.3.5/iris-sklearn
     """
     When the model "overcommit-3" should eventually become Ready with timeout "20s"
-    Then send HTTP inference request with timeout "20s" to model "overcommit-1" with payload:
+    Then I send HTTP inference request with timeout "20s" to model "overcommit-1" with payload:
     """
     {
         "inputs": [
@@ -65,7 +65,7 @@ Feature: Explicit Model deployment
     }
     """
     And expect http response status code "200"
-    Then send HTTP inference request with timeout "20s" to model "overcommit-2" with payload:
+    Then I send HTTP inference request with timeout "20s" to model "overcommit-2" with payload:
     """
     {
         "inputs": [
@@ -79,7 +79,7 @@ Feature: Explicit Model deployment
     }
     """
     And expect http response status code "200"
-    Then send HTTP inference request with timeout "20s" to model "overcommit-3" with payload:
+    Then I send HTTP inference request with timeout "20s" to model "overcommit-3" with payload:
     """
     {
         "inputs": [

tests/integration/godog/features/pipeline/conditional.feature

@@ -1,8 +1,10 @@
-@PipelineDeployment @Functional @Pipelines @Conditional
+@PipelineConditional @Functional @Pipelines @Conditional
 Feature: Conditional pipeline with branching models
-  This pipeline uses a conditional model to route data to either add10 or mul10.
+  In order to support decision-based inference
+  As a model user
+  I need a conditional pipeline that directs inputs to one of multiple models based on a condition
 
-  Scenario: Deploy tfsimple-conditional pipeline and wait for readiness
+  Scenario: Deploy a conditional pipeline, run inference, and verify the output
     Given I deploy model spec with timeout "30s":
     """
     apiVersion: mlops.seldon.io/v1alpha1
@@ -43,7 +45,7 @@ Feature: Conditional pipeline with branching models
     And the model "add10-nbsl" should eventually become Ready with timeout "20s"
     And the model "mul10-nbsl" should eventually become Ready with timeout "20s"
 
-    And I deploy pipeline spec with timeout "30s":
+    When I deploy a pipeline spec with timeout "30s":
     """
     apiVersion: mlops.seldon.io/v1alpha1
     kind: Pipeline
@@ -69,3 +71,70 @@ Feature: Conditional pipeline with branching models
         stepsJoin: any
     """
     Then the pipeline "tfsimple-conditional-nbsl" should eventually become Ready with timeout "40s"
+    Then I send gRPC inference request with timeout "20s" to pipeline "tfsimple-conditional-nbsl" with payload:
+    """
+    {
+      "model_name": "conditional-nbsl",
+      "inputs": [
+        {
+          "name": "CHOICE",
+          "contents": {
+            "int_contents": [
+              0
+            ]
+          },
+          "datatype": "INT32",
+          "shape": [
+            1
+          ]
+        },
+        {
+          "name": "INPUT0",
+          "contents": {
+            "fp32_contents": [
+              1,
+              2,
+              3,
+              4
+            ]
+          },
+          "datatype": "FP32",
+          "shape": [
+            4
+          ]
+        },
+        {
+          "name": "INPUT1",
+          "contents": {
+            "fp32_contents": [
+              1,
+              2,
+              3,
+              4
+            ]
+          },
+          "datatype": "FP32",
+          "shape": [
+            4
+          ]
+        }
+      ]
+    }
+    """
+    And expect gRPC response body to contain JSON:
+    """
+    {
+      "outputs": [
+        {
+          "name": "OUTPUT",
+          "datatype": "FP32",
+          "shape": [
+            4
+          ]
+        }
+      ],
+      "raw_output_contents": [
+        "AAAgQQAAoEEAAPBBAAAgQg=="
+      ]
+    }
+    """