added mcp fallback

backend/routes/search.py

@@ -67,7 +67,7 @@ async def search_math_problems(
         kb_results = await qdrant_service.search_similar(validated_question)
         
         # Step 3: Determine if we need web search fallback
-        confidence_threshold = 0.5
+        confidence_threshold = 0.8  # Increased from 0.5 to 0.8 for higher confidence requirement
         best_score = kb_results[0].score if kb_results else 0.0
         
         if best_score >= confidence_threshold:

backend/services/qdrant_service.py

@@ -32,13 +32,20 @@ class QdrantService:
     def _initialize(self):
         """Initialize Qdrant manager and embedding generator."""
         try:
-            # Qdrant configuration (matching database/ingest.py)
+            import os
+            from dotenv import load_dotenv
+            load_dotenv()
+            
+            # Qdrant configuration from environment variables
             qdrant_config = {
-                'url': 'https://7c49e9a8-f84b-4cc8-9e14-bbffdc2e68ad.us-east4-0.gcp.cloud.qdrant.io:6333',
-                'api_key': 'aFrfsC3xnXVgMEjClC3VNgY2Hgp0f6A5Zd30UM5yQJx4SkEPgn4xSw',
-                'collection_name': 'math_problems'
+                'url': os.getenv('QDRANT_URL'),
+                'api_key': os.getenv('QDRANT_API_KEY'),
+                'collection_name': os.getenv('QDRANT_COLLECTION', 'nuinamath')
             }
             
+            if not qdrant_config['url'] or not qdrant_config['api_key']:
+                raise ValueError("QDRANT_URL and QDRANT_API_KEY must be set in environment variables")
+            
             self.qdrant_manager = QdrantManager(
                 url=qdrant_config['url'],
                 api_key=qdrant_config['api_key']
@@ -68,12 +75,14 @@ class QdrantService:
             return []
         
         try:
+            import os
             # Generate embedding for the question
             query_embedding = self.embedding_generator.embed_text(question)
             
             # Search in Qdrant
+            collection_name = os.getenv('QDRANT_COLLECTION', 'nuinamath')
             results = self.qdrant_manager.search_similar(
-                collection_name='math_problems',
+                collection_name=collection_name,
                 query_vector=query_embedding,
                 limit=limit
             )
@@ -132,7 +141,8 @@ class QdrantService:
                 request_data=request_data,
                 response_data=response_data,
                 response_time_ms=response_time_ms,
-                source=source
+                source=source,
+                status_code=200  # Default to 200 for successful responses
             )
             
             # TODO: Store log entry in Qdrant analytics collection

database/.gitignore

@@ -0,0 +1,23 @@
+# Python cache
+__pycache__/
+*.py[cod]
+*$py.class
+
+# Virtual environment
+venv/
+env/
+
+# IDE files
+.vscode/
+.idea/
+
+# OS files
+.DS_Store
+Thumbs.db
+
+# Logs
+*.log
+
+# Temporary files
+*.tmp
+*.temp

database/test_retrieval.py

@@ -1,93 +0,0 @@
-"""
-Test script for retrieving similar math problems from Qdrant.
-"""
-import logging
-import os
-from dotenv import load_dotenv
-
-# Load environment variables
-load_dotenv()
-
-# Configuration settings
-QDRANT_URL = os.getenv("QDRANT_URL")
-QDRANT_API_KEY = os.getenv("QDRANT_API_KEY") 
-QDRANT_COLLECTION = os.getenv("QDRANT_COLLECTION", "nuinamath")
-EMBEDDING_MODEL = "all-MiniLM-L6-v2"
-
-from utils import EmbeddingGenerator, format_retrieval_results
-from qdrant_manager import QdrantManager
-
-# Set up logging
-logging.basicConfig(level=logging.INFO)
-logger = logging.getLogger(__name__)
-
-def test_retrieval():
-    """Test the retrieval system with sample math questions."""
-    
-    # Sample test questions
-    test_questions = [
-        "What is the value of x in 3x + 5 = 20?",
-        "How do you find the area of a triangle given 3 sides?",
-        "Solve for y: 2y - 7 = 15",
-        "What is the derivative of x^2 + 3x?",
-        "Find the arithmetic sequence common difference"
-    ]
-    
-    try:
-        # Initialize components
-        logger.info("Initializing retrieval system...")
-        embedding_generator = EmbeddingGenerator(EMBEDDING_MODEL)
-        qdrant_manager = QdrantManager(QDRANT_URL, QDRANT_API_KEY)
-        
-        # Test each question
-        for i, question in enumerate(test_questions, 1):
-            print(f"\n{'='*60}")
-            print(f"TEST QUERY {i}: {question}")
-            print('='*60)
-            
-            # Generate embedding for the question
-            query_embedding = embedding_generator.embed_single_text(question)
-            
-            # Search for similar problems
-            results = qdrant_manager.search_similar(
-                collection_name=QDRANT_COLLECTION,
-                query_vector=query_embedding,
-                limit=3,
-                score_threshold=0.1
-            )
-            
-            # Format and display results
-            formatted_results = format_retrieval_results(results)
-            print(formatted_results)
-    
-    except Exception as e:
-        logger.error(f"Error in retrieval test: {e}")
-
-def test_collection_status():
-    """Check the status of the Qdrant collection."""
-    try:
-        qdrant_manager = QdrantManager(QDRANT_URL, QDRANT_API_KEY)
-        
-        print(f"\n{'='*40}")
-        print("COLLECTION STATUS")
-        print('='*40)
-        
-        info = qdrant_manager.get_collection_info(QDRANT_COLLECTION)
-        if info:
-            print(f"Collection Name: {QDRANT_COLLECTION}")
-            print(f"Status: {info.status}")
-            print(f"Vectors Count: {info.vectors_count}")
-            print(f"Vector Size: {info.config.params.vectors.size}")
-            print(f"Distance Metric: {info.config.params.vectors.distance}")
-        else:
-            print("Collection not found or error occurred")
-            
-    except Exception as e:
-        logger.error(f"Error checking collection status: {e}")
-
-if __name__ == "__main__":
-    print("Testing Qdrant Collection Status...")
-    test_collection_status()
-    
-    print("\n\nTesting Retrieval System...")
-    test_retrieval()