diff --git a/platypush/plugins/assistant/picovoice/__init__.py b/platypush/plugins/assistant/picovoice/__init__.py
index c669f19f9..9426ba608 100644
--- a/platypush/plugins/assistant/picovoice/__init__.py
+++ b/platypush/plugins/assistant/picovoice/__init__.py
@@ -40,6 +40,10 @@ class AssistantPicovoicePlugin(AssistantPlugin, RunnablePlugin):
         * **Leopard**: speech-to-text engine aimed at offline transcription of
           audio files rather than real-time transcription.
 
+        * **Orca**: text-to-speech engine, if you want to create your custom
+          logic to respond to user's voice commands and render the responses as
+          audio.
+
     """
 
     def __init__(
diff --git a/platypush/plugins/assistant/picovoice/_assistant.py b/platypush/plugins/assistant/picovoice/_assistant.py
index e511bc01c..25142f336 100644
--- a/platypush/plugins/assistant/picovoice/_assistant.py
+++ b/platypush/plugins/assistant/picovoice/_assistant.py
@@ -287,10 +287,8 @@ class Assistant:
             if self.cheetah:
                 phrase = self.cheetah.flush()
 
-            if not self._ctx.is_final:
-                self._ctx.partial_transcript += phrase
-                phrase = self._ctx.partial_transcript
-
+            self._ctx.partial_transcript += phrase
+            phrase = self._ctx.partial_transcript
             phrase = phrase[:1].lower() + phrase[1:]
 
             if self._ctx.is_final or phrase: