diff --git a/services/hermes/voice-deployment.yaml b/services/hermes/voice-deployment.yaml
index 1f249032b..e1f8bf2e0 100644
--- a/services/hermes/voice-deployment.yaml
+++ b/services/hermes/voice-deployment.yaml
@@ -25,6 +25,8 @@ spec:
       runtimeClassName: nvidia
       automountServiceAccountToken: false
       enableServiceLinks: false
+      securityContext:
+        supplementalGroups: [44]
       nodeSelector:
         kubernetes.io/hostname: titan-21
       containers:
diff --git a/testing/tests/test_hermes_chat_quality.py b/testing/tests/test_hermes_chat_quality.py
index 9f65f105c..e0259c8dd 100644
--- a/testing/tests/test_hermes_chat_quality.py
+++ b/testing/tests/test_hermes_chat_quality.py
@@ -232,6 +232,7 @@ def test_voice_workloads_have_deliberate_xavier_placement():
     assert stt["enableServiceLinks"] is False
     assert tts["enableServiceLinks"] is False
     assert stt["runtimeClassName"] == "nvidia"
+    assert stt["securityContext"]["supplementalGroups"] == [44]
     stt_resources = stt["containers"][0]["resources"]
     assert stt_resources["requests"]["nvidia.com/gpu.shared"] == 1
     assert stt_resources["limits"]["nvidia.com/gpu.shared"] == 1