Merge branch 'main' of https://github.com/parvbhullar/unpod

# Conflicts:
#	.gitignore
#	apps/backend-core/config/settings/base.py
#	apps/backend-core/unpod/common/services/livekit.py
#	apps/backend-core/unpod/common/utils.py
#	apps/backend-core/unpod/core_components/reports/daily.py
#	apps/backend-core/unpod/core_components/serializers.py
#	apps/backend-core/unpod/core_components/services.py
#	apps/backend-core/unpod/core_components/tasks/utils.py
#	apps/backend-core/unpod/core_components/views.py
#	apps/backend-core/unpod/users/tests/test_forms.py
#	apps/super/Dockerfile
#	apps/super/deployment/cerebrium/Dockerfile
#	apps/super/deployment/cerebrium/cerebrium.toml
#	apps/super/deployment/cerebrium/setup.py
#	apps/super/deployment/modal/modal_app.py
#	apps/super/deployment/modal/setup.py
#	apps/super/pyproject.toml
#	apps/super/requirements/super.txt
#	apps/super/requirements/super_services.txt
#	apps/super/scripts/test_voice_agent_coverage.sh
#	apps/super/scripts/test_voice_agent_suite.sh
#	apps/super/super/app/call_execution.py
#	apps/super/super/app/providers/__init__.py
#	apps/super/super/app/providers/factory.py
#	apps/super/super/core/voice/livekit/event_bridge.py
#	apps/super/super/core/voice/livekit/lite_handler.py
#	apps/super/super/core/voice/livekit/livekit_lite_agent.py
#	apps/super/super/core/voice/livekit/state_observer.py
#	apps/super/super/core/voice/pipecat/lite_handler.py
#	apps/super/super/core/voice/pipecat/services.py
#	apps/super/super/core/voice/plugins/transcript_plugin.py
#	apps/super/super/core/voice/providers/base.py
#	apps/super/super/core/voice/services/lazy_factory.py
#	apps/super/super/core/voice/services/livekit_services.py
#	apps/super/super/core/voice/voice_agent_handler.py
#	apps/super/super/pyproject.toml
#	apps/super/super_services/evals/eval_generator.py
#	apps/super/super_services/libs/config.py
#	apps/super/super_services/orchestration/cron_jobs/eval_testing.py
#	apps/super/super_services/orchestration/cron_jobs/post_call.py
#	apps/super/super_services/orchestration/executors/voice_executor_v3.py
#	apps/super/super_services/orchestration/task/deployments/requirements.txt
#	apps/super/super_services/orchestration/task/task_service.py
#	apps/super/super_services/voice/consumers/prefect_task_consumer.py
#	apps/super/tests/conftest.py
#	apps/super/tests/core/voice/helpers/runtime_imports.py
#	apps/super/tests/core/voice/livekit/test_lite_handler_cleanup.py
#	apps/super/tests/core/voice/services/test_livekit_config_parity.py
#	apps/super/tests/core/voice/services/test_pipecat_inference.py
#	apps/super/tests/core/voice/test_kb_chroma_kn_base_integration.py
#	apps/super/tests/core/voice/test_voice_agent_suite_manifest.py

Author: divyanshu-khudiya

.gitignore

@@ -61,6 +61,7 @@ next-env.d.ts
 # Environment
 .env
 .env.*
+*.env
 !.env.example
 !.env.*.example
 .env.local

apps/backend-core/unpod/common/services/livekit.py

@@ -1,10 +1,3 @@
-import time
-
-import jwt
-import requests
-from django.conf import settings
-
-
 def get_livekit_trunks():
   url = f"https://{settings.LIVEKIT_BASE}/twirp/livekit.SIP/ListSIPInboundTrunk"
 

apps/backend-core/unpod/core_components/reports/daily.py

@@ -86,7 +86,7 @@ def get_data_mongo(collection, today, query: dict = {}):
     )
 
 
-def get_org_wise_data_mongo(collection, today, query: dict = {}):
+def get_hubwise_data_mongo(collection, today, query: dict = {}):
     week_date = today - timezone.timedelta(days=7)
     month_date = today - timezone.timedelta(days=30)
     last_day = today - timezone.timedelta(days=1)
@@ -141,31 +141,31 @@ def get_org_wise_data_mongo(collection, today, query: dict = {}):
     }
 
 
-def merge_organization_data(orgDataDictId, space_data, post_data):
+def merge_hub_data(organizationDictId, space_data, post_data):
     merged_data = {}
-    for organization in orgDataDictId:
-        merge_org_data = []
+    for organization in organizationDictId:
+        merge_hub_data = []
         stats_keys = ["total", "today", "last_day", "weekly", "monthly"]
         for stats in stats_keys:
-            merge_org_data.append(space_data[stats].get(organization, 0))
-            merge_org_data.append(post_data[stats].get(organization, 0))
-        merged_data[orgDataDictId[organization]["name"]] = merge_org_data
+            merge_hub_data.append(space_data[stats].get(organization, 0))
+            merge_hub_data.append(post_data[stats].get(organization, 0))
+        merged_data[organizationDictId[organization]["name"]] = merge_hub_data
     return merged_data
 
 
-def orgWiseData(organizationList, today=None):
+def hubWiseData(hubData, today=None):
     if today is None:
         today = timezone.now().date().today()
-    orgDataDictId = {}
-    orgDataDictHandle = {}
-    for each in organizationList:
-        orgDataDictHandle[each["domain_handle"]] = each
-        orgDataDictId[each["id"]] = each
+    hubDataDictId = {}
+    hubDataDictHandle = {}
+    for each in hubData:
+        hubDataDictHandle[each["domain_handle"]] = each
+        hubDataDictId[each["id"]] = each
     space_data = Space.objects.all()
     post_data = ThreadPost.objects.all()
-    space_data = get_organization_group_data(space_data, today, "space_organization_id")
-    post_data = get_organization_group_data(post_data, today, "space__space_organization_id")
-    merged_data = merge_organization_data(orgDataDictId, space_data, post_data)
+    space_data = get_hub_group_data(space_data, today, "space_organization_id")
+    post_data = get_hub_group_data(post_data, today, "space__space_organization_id")
+    merged_data = merge_hub_data(hubDataDictId, space_data, post_data)
     return merged_data
 
 
@@ -210,7 +210,7 @@ def dailyActivityReport():
     body += "<br><br>"
     body += """<h3 style="text-align:center; margin-top:10px;">Last 30 Days Organization Wise Data</h3>"""
     organizationList = SpaceOrganization.objects.all().values("id", "domain_handle", "name")
-    organizationData = orgWiseData(organizationList, today)
+    organizationData = hubWiseData(organizationList, today)
     body += """
             <table width="100%" border="2" cellspacing="10" cellpadding="10" rules="ALL" style="border-collapse:collapse;color:#1f2240;background-color:#fff;padding:10px">
                 <tr>
@@ -237,7 +237,7 @@ def dailyActivityReport():
             """
     organizationWisePilotData = {}
     for organization in organizationList:
-        pilot_data = get_org_wise_data_mongo(
+        pilot_data = get_hubwise_data_mongo(
             "task_request_log", today, {"org_id": organization["id"]}
         )
         organizationWisePilotData[organization["name"]] = pilot_data

apps/backend-core/unpod/core_components/serializers.py

@@ -413,7 +413,6 @@ class PilotFullSerializer(serializers.ModelSerializer):
     template = serializers.CharField(source="template.slug", read_only=True)
     eval_kn_bases = serializers.SerializerMethodField()
     eval_knowledge_bases = serializers.SerializerMethodField()
-    pilot_eval_info = serializers.SerializerMethodField()
 
     class Meta:
         model = Pilot
@@ -481,7 +480,6 @@ class Meta:
             "realtime_evals",
             "eval_kn_bases",
             "eval_knowledge_bases",
-            "pilot_eval_info",
             "components",
         )
 
@@ -582,29 +580,6 @@ def get_eval_knowledge_bases(self, obj):
             obj.eval_kn_bases.all().values("name", "slug", "token")
         )
 
-    def get_pilot_eval_info(self, obj):
-        from unpod.knowledge_base.models import KnowledgeBaseEvals
-        eval_info = KnowledgeBaseEvals.objects.filter(
-            linked_handle=obj.handle, eval_type="pilot"
-        ).first()
-        eval_data = {}
-        if eval_info:
-            eval_data = {
-                "eval_name": eval_info.eval_name,
-                "eval_type": eval_info.eval_type,
-                "linked_handle": eval_info.linked_handle,
-                "gen_status": eval_info.gen_status,
-            }
-            if eval_info.eval_data:
-                eval_data["eval_token"] = eval_info.eval_data.get("space_token")
-        if eval_data.get("eval_token"):
-            eval_slug = Space.objects.filter(
-                token=eval_data["eval_token"], space_type="knowledge_base"
-            ).values("slug").first()
-            if eval_slug:
-                eval_data["eval_slug"] = eval_slug["slug"]
-        return eval_data
-
     def get_organization(self, obj):
         if hasattr(obj, "owner"):
             return SpaceOrganizationSerializers(obj.owner).data

apps/backend-core/unpod/core_components/services.py

@@ -371,40 +371,32 @@ def process_pilot_knowledge_bases(pilot, knowledge_bases, is_update=False):
         if knowledge_bases is None:
             return
 
-        total_req_kb = len(knowledge_bases)
+        # Validate all knowledge bases exist
+        kb_count = Space.objects.filter(
+            slug__in=knowledge_bases, space_type=SpaceType.knowledge_base.name
+        ).count()
+
+        if kb_count != len(knowledge_bases):
+            raise APIException206(detail={"message": "Knowledge Bases Not Found"})
+
+        kbs = Space.objects.filter(
+            slug__in=knowledge_bases, space_type=SpaceType.knowledge_base.name
+        )
+
         content_type = ContentType.objects.get_for_model(Space)
 
-        if total_req_kb == 0 or (total_req_kb == 1 and knowledge_bases[0] == "--"):
+        # For updates, delete existing KB links
+        if is_update:
             PilotLink.objects.filter(
                 pilot=pilot,
                 content_type=content_type,
                 spaces__space_type=SpaceType.knowledge_base.name,
             ).delete()
-        else:
-            # Validate all knowledge bases exist
-            kb_count = Space.objects.filter(
-                slug__in=knowledge_bases, space_type=SpaceType.knowledge_base.name
-            ).count()
-
-            if kb_count != total_req_kb:
-                raise APIException206(detail={"message": "Knowledge Bases Not Found"})
-
-            kbs = Space.objects.filter(
-                slug__in=knowledge_bases, space_type=SpaceType.knowledge_base.name
-            )
 
-            # For updates, delete existing KB links
-            if is_update:
-                PilotLink.objects.filter(
-                    pilot=pilot,
-                    content_type=content_type,
-                    spaces__space_type=SpaceType.knowledge_base.name,
-                ).delete()
-
-            # Bulk create new KB links
-            PilotLink.objects.bulk_create(
-                [PilotLink(pilot=pilot, content_object=kb) for kb in kbs]
-            )
+        # Bulk create new KB links
+        PilotLink.objects.bulk_create(
+            [PilotLink(pilot=pilot, content_object=kb) for kb in kbs]
+        )
 
     @staticmethod
     def process_pilot_components(pilot, components, request):

apps/backend-core/unpod/users/tests/test_forms.py

@@ -4,15 +4,15 @@
 import pytest
 from django.utils.translation import gettext_lazy as _
 
-from unpod.users.forms import UserAdminCreationForm
+from unpod.users.forms import UserCreationForm
 from unpod.users.models import User
 
 pytestmark = pytest.mark.django_db
 
 
-class TestUserAdminCreationForm:
+class TestUserCreationForm:
     """
-    Test class for all tests related to the UserAdminCreationForm
+    Test class for all tests related to the UserCreationForm
     """
 
     def test_username_validation_error_msg(self, user: User):
@@ -25,7 +25,7 @@ def test_username_validation_error_msg(self, user: User):
 
         # The user already exists,
         # hence cannot be created.
-        form = UserAdminCreationForm(
+        form = UserCreationForm(
             {
                 "username": user.username,
                 "password1": user.password,

apps/super/Dockerfile

@@ -64,6 +64,7 @@ COPY --from=builder /app /app
 ENV SETTINGS_FILE="super_services.settings.qa"
 ENV PATH="/app/.venv/bin:$PATH"
 RUN SKIP_DB_CHECK=1 python super_services/orchestration/executors/voice_executor_v3.py download-files
+RUN python -c "from chromadb.utils.embedding_functions import ONNXMiniLM_L6_V2; ONNXMiniLM_L6_V2()"
 
 # Switch to non-privileged user
 RUN chown -R appuser:appuser /app

apps/super/deployment/cerebrium/Dockerfile

@@ -21,6 +21,8 @@ COPY __init__.py .
 ENV PYTHONPATH="/app"
 ENV PYTHONUNBUFFERED=1
 ENV SETTINGS_FILE="super_services.settings.prod"
+# Pin HuggingFace cache inside /app so models survive build→runtime
+ENV HF_HOME="/app/.hf_cache"
 
 # Pre-download ML models at build time (SKIP_DB_CHECK: no DB during build)
 RUN SKIP_DB_CHECK=1 python super_services/orchestration/executors/voice_executor_v3.py download-files

apps/super/deployment/cerebrium/cerebrium.toml

@@ -32,16 +32,41 @@ exclude = [
 cpu = 4
 memory = 8.0
 compute = "CPU"
+provider = "aws"
+region = "us-east-1"
 
 [cerebrium.scaling]
 min_replicas = 1
 max_replicas = 5
 cooldown = 30
 replica_concurrency = 1
 
+[cerebrium.dependencies.apt]
+gcc = "latest"
+"g++" = "latest"
+python3-dev = "latest"
+libpq-dev = "latest"
+git = "latest"
+curl = "latest"
+portaudio19-dev = "latest"
+ffmpeg = "latest"
+libsndfile1 = "latest"
+
 [cerebrium.dependencies.paths]
 pip = "requirements.txt"
 
+[cerebrium.environment]
+PYTHONPATH = "/cortex"
+PYTHONUNBUFFERED = "1"
+SETTINGS_FILE = "super_services.settings.prod"
+HF_HOME = "/cortex/.hf_cache"
+
+[cerebrium.build]
+shell_commands = [
+    "SKIP_DB_CHECK=1 python /cortex/super_services/orchestration/executors/voice_executor_v3.py download-files",
+]
+
 [cerebrium.runtime.custom]
+entrypoint = ["python", "-u", "/cortex/super_services/orchestration/executors/voice_executor_v3.py", "start"]
 port = 8600
-dockerfile_path = "./deployment/cerebrium/Dockerfile"
+healthcheck_endpoint = "/"