Update app.py

app.py

@@ -16,7 +16,6 @@ import matplotlib.pyplot as plt
 from huggingface_hub import login
 import os
 from dotenv import load_dotenv
-import shutil
 
 SPREADSHEET_ID = "1CsBub3Jlwyo7WHMQty6SDnBShIZMjl5XTVSoOKrxZhc"
 RANGE_NAME = 'Sheet1!A1:E'
@@ -26,11 +25,6 @@ SERVICE_ACCOUNT_FILE = r"C:\Users\bhagy\AI\credentials.json"
 csv_file_path = r"C:\Users\bhagy\OneDrive\Desktop\INFOSYS PROJECT\900_products_dataset.csv"
 
 
-# storage_path = "chromadb_storage"
-# if not os.path.exists(storage_path):
-#     os.makedirs(storage_path)
-
-# Define Custom Embedding Function
 class CustomEmbeddingFunction:
     def __init__(self, model_name="sentence-transformers/all-MiniLM-L6-v2"):
         self.tokenizer = AutoTokenizer.from_pretrained(model_name)
@@ -43,37 +37,15 @@ class CustomEmbeddingFunction:
         embeddings = outputs.last_hidden_state.mean(dim=1).squeeze().numpy()
         return embeddings
 
-# Initialize Sentiment Analysis Pipeline
+# Initialize components
 sentiment_pipeline = pipeline("sentiment-analysis")
-
-# Ensure no corrupted ChromaDB storage
-storage_path = r"C:\Users\bhagy\OneDrive\Desktop\INFOSYS PROJECT\chromadb_storage" # Replace with the correct path
-
-try:
-    # Initialize ChromaDB Client
-    chroma_client = Client(Settings(persist_directory=storage_path))
-except ValueError as e:
-    print(f"Error: {str(e)}")
-    print("ChromaDB storage may be corrupted or the tenant might be missing. Resetting storage...")
-    
-    # Remove the existing storage path and recreate it
-    shutil.rmtree(storage_path, ignore_errors=True)
-    os.makedirs(storage_path)
-    
-    # Try reconnecting to ChromaDB after reset
-    chroma_client = Client(Settings(persist_directory=storage_path))
-
-# Initialize Embedding Function
+chroma_client = Client(Settings(persist_directory="chromadb_storage"))
 embedding_fn = CustomEmbeddingFunction()
-
-# Define collection name
 collection_name = "crm_data"
 
-# Try fetching collection; create if missing
 try:
     collection = chroma_client.get_collection(collection_name)
-except Exception as e:
-    print(f"Error fetching collection: {str(e)}. Creating new collection...")
+except Exception:
     collection = chroma_client.create_collection(collection_name)
 
 def get_google_sheets_service():
@@ -111,11 +83,8 @@ def update_google_sheet(transcribed_text, sentiment,objection, recommendations,o
         st.error(f"Failed to update Google Sheets: {e}")
 
 load_dotenv()
-hf_token= os.getenv("HUGGINGFACE_TOKEN")
-login(token=hf_token)
-if not hf_token:
-    raise ValueError("Hugging Face API key not found! Please set the HUGGINGFACE_TOKEN variable.")
-print(f"API Key Loaded: {hf_token[:5]}****")
+huggingface_api_key= os.getenv("HUGGINGFACE_TOKEN")
+login(token=huggingface_api_key)
 
 model_name = "tabularisai/multilingual-sentiment-analysis"
 model = AutoModelForSequenceClassification.from_pretrained(model_name)
@@ -134,7 +103,7 @@ def analyze_sentiment(text):
         
         print(f"Sentiment Analysis Result: {result}")
         
-        
+        # Map raw labels to sentiments
         sentiment_map = {
             'Very Negative': "NEGATIVE",
             'Negative': "NEGATIVE",
@@ -208,28 +177,9 @@ def query_crm_data_with_context(prompt, top_k=3):
 
 
 
-
-storage_path = r"C:\Users\bhagy\OneDrive\Desktop\INFOSYS PROJECT\chromadb_storage"  # Update with your storage path
-
-# Ensure the storage path exists
-if not os.path.exists(storage_path):
-    print(f"Storage path {storage_path} does not exist. Creating it...")
-    os.makedirs(storage_path)
-
-# Clear any corrupted data
-if os.path.exists(storage_path):
-    print(f"Removing existing storage at {storage_path}...")
-    shutil.rmtree(storage_path)
-    os.makedirs(storage_path)
-
-# Initialize the sentence transformer model
 sentence_model = SentenceTransformer('sentence-transformers/all-MiniLM-L6-v2')
+faiss_index = faiss.IndexFlatL2(384)
 
-# Create Faiss index with the correct dimensionality (make sure this matches the embedding size)
-expected_dim = 384  # Example value for sentence embeddings
-faiss_index = faiss.IndexFlatL2(expected_dim)
-
-# Load objection responses
 def load_objection_responses(csv_file_path):
     try:
         df = pd.read_csv(csv_file_path)
@@ -239,31 +189,10 @@ def load_objection_responses(csv_file_path):
         print(f"Error loading objections CSV: {e}")
         return {}
 
-# Load objections and encode them into embeddings
 objection_response_pairs = load_objection_responses(r"C:\Users\bhagy\OneDrive\Desktop\INFOSYS PROJECT\objections_responses.csv")
 objections = list(objection_response_pairs.keys())
 objection_embeddings = sentence_model.encode(objections)
-objection_embeddings = np.array(objection_embeddings, dtype="float32")
-
-# Check the shape of the embeddings
-print(f"Shape of objection_embeddings: {objection_embeddings.shape}")
-
-# Check if the embeddings dimensionality matches the expected dimension
-# If the shape is (1,), reshape it to (1, expected_dim)
-if len(objection_embeddings.shape) == 1:
-    # Reshape for a single embedding, turn it into a 2D array
-    objection_embeddings = objection_embeddings.reshape(1, -1)
-elif len(objection_embeddings.shape) == 2:
-    # Check if second dimension matches expected_dim
-    if objection_embeddings.shape[1] != expected_dim:
-        raise ValueError(f"Dimensionality of embeddings {objection_embeddings.shape[1]} does not match expected dimension {expected_dim}.")
-else:
-    raise ValueError(f"Unexpected shape for objection embeddings: {objection_embeddings.shape}")
-
-# Add the embeddings to the Faiss index
-faiss_index.add(objection_embeddings)
-
-print(f"Successfully added {objection_embeddings.shape[0]} embeddings to the Faiss index.")
+faiss_index.add(np.array(objection_embeddings, dtype="float32"))
 
 def find_closest_objection(query):
     query_embedding = sentence_model.encode([query])
@@ -609,8 +538,6 @@ def generate_post_call_summary(sentiment_history, recommendations=[]):
 
 # Main
 def main():
-    
-    st.set_page_config(page_title="RealTime AI-Powered Sales Assistant", layout="wide")
     st.title("🤖 RealTime AI-Powered Sales Assistant For Enhanced Conversation")
     st.markdown(
         "An intelligent assistant to analyze speech, handle objections, and recommend products in real-time."