diff --git a/tools/cloudflow-blueprint/src/main/scala/cloudflow/blueprint/deployment/ApplicationDescriptor.scala b/tools/cloudflow-blueprint/src/main/scala/cloudflow/blueprint/deployment/ApplicationDescriptor.scala
index 95d8399ed..dda79e825 100644
--- a/tools/cloudflow-blueprint/src/main/scala/cloudflow/blueprint/deployment/ApplicationDescriptor.scala
+++ b/tools/cloudflow-blueprint/src/main/scala/cloudflow/blueprint/deployment/ApplicationDescriptor.scala
@@ -65,7 +65,8 @@ object ApplicationDescriptor {
         }
 
     ApplicationDescriptor(sanitizedApplicationId, appVersion, namedStreamletDescriptors.map {
-      case (_, instance) => instance
+      case (_, instance) =>
+        StreamletInstance(instance.name, sanitizeDescriptor(instance.descriptor))
     }, deployments, agentPaths, Version, libraryVersion)
   }
 
@@ -77,6 +78,17 @@ object ApplicationDescriptor {
     }.toMap
   private def streamletToNamedStreamletDescriptor(streamlet: VerifiedStreamlet) =
     (streamlet, StreamletInstance(streamlet.name, streamlet.descriptor))
+
+  /**
+   *  Deletes every schema
+   *  StreamletDescriptor.[inlets | outlets].SchemaDescriptor.schema
+   *  to avoid adding the description of each type of the schema in the CR
+   */
+  private def sanitizeDescriptor(descriptor: StreamletDescriptor): StreamletDescriptor = {
+    val sanitizedInlets = descriptor.inlets.map(each => each.copy(schema = each.schema.copy(schema = "")))
+    val sanitizedOutlets = descriptor.outlets.map(each => each.copy(schema = each.schema.copy(schema = "")))
+    descriptor.copy(inlets = sanitizedInlets, outlets = sanitizedOutlets)
+  }
 }
 
 /**
diff --git a/tools/cloudflow-blueprint/src/test/scala/cloudflow/blueprint/deployment/ApplicationDescriptorSpec.scala b/tools/cloudflow-blueprint/src/test/scala/cloudflow/blueprint/deployment/ApplicationDescriptorSpec.scala
index 1da242c52..004ada434 100644
--- a/tools/cloudflow-blueprint/src/test/scala/cloudflow/blueprint/deployment/ApplicationDescriptorSpec.scala
+++ b/tools/cloudflow-blueprint/src/test/scala/cloudflow/blueprint/deployment/ApplicationDescriptorSpec.scala
@@ -214,6 +214,44 @@ class ApplicationDescriptorSpec
       processor2Deployment.portMappings.map { case (port, sp) => port -> sp.id } must contain("out" -> "foos2")
     }
 
+    "from a verified blueprint clean the schema from the CR" in {
+      Given("a verified blueprint")
+      val ingress = randomStreamlet().asIngress[Foo].withServerAttribute
+      val processor = randomStreamlet().asProcessor[Foo, Bar].withRuntime("spark")
+      val egress = randomStreamlet().asEgress[Bar]
+
+      val ingressRef = ingress.ref("ingress")
+      val processorRef = processor.ref("processor")
+      val egressRef = egress.ref("egress")
+
+      val verifiedBlueprint = Blueprint()
+        .define(Vector(ingress, processor, egress))
+        .use(ingressRef)
+        .use(processorRef)
+        .use(egressRef)
+        .connect(BTopic("foos"), ingressRef.out, processorRef.in)
+        .connect(BTopic("bars1"), processorRef.out, egressRef.in)
+        .verified
+        .right
+        .value
+
+      When("I create a deployment descriptor from that blueprint")
+      val appId = "noisy-nissan-42"
+      val appVersion = "1-2345678"
+      val image = "image-1"
+      val descriptor = ApplicationDescriptor(appId, appVersion, image, verifiedBlueprint, agentPaths, BuildInfo.version)
+
+      Then("the descriptor must be valid")
+      descriptor.deployments.size mustBe 3
+      Then("the description of the streamlets should not contain the SchemaDescriptor.schema populated")
+      val schemas = for {
+        streamlet <- descriptor.streamlets
+        port <- streamlet.descriptor.inlets ++ streamlet.descriptor.outlets
+      } yield port.schema.schema
+      schemas mustBe Vector("", "", "", "")
+
+    }
+
     "be built correctly from a verified blueprint (with dual-inlet merging)" in {
       Given("a verified blueprint")
       val ingress1 = randomStreamlet().asIngress[Foo].withServerAttribute