- Adding a Tenant Type

- Allow filtering on Tenant Types & searching for parts of Tenant names
- Implement health checks
- Start Prometheus monitoring (needs to be finalized)
- Refine audio_processor and srt_processor to reduce duplicate code and support for larger files
- Introduce repopack to reason in LLMs about the code

common/extensions.py

@@ -10,8 +10,8 @@ from flask_jwt_extended import JWTManager
 from flask_session import Session
 from flask_wtf import CSRFProtect
 from flask_restx import Api
+from prometheus_flask_exporter import PrometheusMetrics
 
-from .utils.nginx_utils import prefixed_url_for
 from .utils.simple_encryption import SimpleEncryption
 from .utils.minio_utils import MinioClient
 
@@ -31,3 +31,4 @@ session = Session()
 api_rest = Api()
 simple_encryption = SimpleEncryption()
 minio_client = MinioClient()
+metrics = PrometheusMetrics.for_app_factory()

common/models/user.py

@@ -21,6 +21,7 @@ class Tenant(db.Model):
     website = db.Column(db.String(255), nullable=True)
     timezone = db.Column(db.String(50), nullable=True, default='UTC')
     rag_context = db.Column(db.Text, nullable=True)
+    type = db.Column(db.String(20), nullable=True, server_default='Active')
 
     # language information
     default_language = db.Column(db.String(2), nullable=True)
@@ -56,7 +57,6 @@ class Tenant(db.Model):
     encrypted_chat_api_key = db.Column(db.String(500), nullable=True)
     encrypted_api_key = db.Column(db.String(500), nullable=True)
 
-
     # Tuning enablers
     embed_tuning = db.Column(db.Boolean, nullable=True, default=False)
     rag_tuning = db.Column(db.Boolean, nullable=True, default=False)
@@ -75,6 +75,7 @@ class Tenant(db.Model):
             'website': self.website,
             'timezone': self.timezone,
             'rag_context': self.rag_context,
+            'type': self.type,
             'default_language': self.default_language,
             'allowed_languages': self.allowed_languages,
             'embedding_model': self.embedding_model,

common/utils/model_utils.py

@@ -147,10 +147,10 @@ def select_model_variables(tenant):
             match llm_model:
                 case 'gpt-4o' | 'gpt-4o-mini':
                     tool_calling_supported = True
-                    PDF_chunk_size = 10000
-                    PDF_chunk_overlap = 200
-                    PDF_min_chunk_size = 8000
-                    PDF_max_chunk_size = 12000
+                    processing_chunk_size = 10000
+                    processing_chunk_overlap = 200
+                    processing_min_chunk_size = 8000
+                    processing_max_chunk_size = 12000
                 case _:
                     raise Exception(f'Error setting model variables for tenant {tenant.id} '
                                     f'error: Invalid chat model')
@@ -165,18 +165,18 @@ def select_model_variables(tenant):
                                                           model=llm_model_ext,
                                                           temperature=model_variables['RAG_temperature'])
             tool_calling_supported = True
-            PDF_chunk_size = 10000
-            PDF_chunk_overlap = 200
-            PDF_min_chunk_size = 8000
-            PDF_max_chunk_size = 12000
+            processing_chunk_size = 10000
+            processing_chunk_overlap = 200
+            processing_min_chunk_size = 8000
+            processing_max_chunk_size = 12000
         case _:
             raise Exception(f'Error setting model variables for tenant {tenant.id} '
                             f'error: Invalid chat provider')
 
-    model_variables['PDF_chunk_size'] = PDF_chunk_size
-    model_variables['PDF_chunk_overlap'] = PDF_chunk_overlap
-    model_variables['PDF_min_chunk_size'] = PDF_min_chunk_size
-    model_variables['PDF_max_chunk_size'] = PDF_max_chunk_size
+    model_variables['processing_chunk_size'] = processing_chunk_size
+    model_variables['processing_chunk_overlap'] = processing_chunk_overlap
+    model_variables['processing_min_chunk_size'] = processing_min_chunk_size
+    model_variables['processing_max_chunk_size'] = processing_max_chunk_size
 
     if tool_calling_supported:
         model_variables['cited_answer_cls'] = CitedAnswer