diff --git a/kork-artifacts/src/main/java/com/netflix/spinnaker/kork/artifacts/README.md b/kork-artifacts/src/main/java/com/netflix/spinnaker/kork/artifacts/README.md
index 5cbd83af1..e838fea49 100644
--- a/kork-artifacts/src/main/java/com/netflix/spinnaker/kork/artifacts/README.md
+++ b/kork-artifacts/src/main/java/com/netflix/spinnaker/kork/artifacts/README.md
@@ -44,10 +44,9 @@ When bootstrapping Spring we add in custom bean serializers and deserializers to
 handle storage or retrieval of an artifact.
 
 Rosco is primarily used for baking artifacts which will generate something
-deployable. When Spring sends a request back to Orca for a bake request, the
-custom serializer injected at Orca startup, will use the artifact storage and
-store the artifact, and instead return a `remote/base64` artifact instead of the
-usual `embedded/base64`.
+deployable.  When Rosco responds to a bake request, the custom serializer
+injected in Rosco at startup stores the artifact and returns a `remote/base64`
+artifact instead of the usual `embedded/base64`.
 
 Clouddriver, for this document, handles mostly deployment and has some endpoints
 regarding artifacts. It does do a little more than this, but we only care about