Update app.py

app.py

@@ -21,7 +21,7 @@ class MultiClientThemeClassifier:
         if not self.model_loaded:
             try:
                 # Using a smaller, faster model for demo
-                self.model = SentenceTransformer('all-MiniLM-L6-v2')
+                self.model = SentenceTransformer('jinaai/jina-embeddings-v3')
                 self.model_loaded = True
                 return "✅ Model loaded successfully!"
             except Exception as e:
@@ -210,9 +210,9 @@ def benchmark_interface(csv_file, client_id: str):
         return f"❌ Error reading CSV: {str(e)}", "", ""
 
 # Create the Gradio interface
-with gr.Blocks(title="Lucidya Custom Themes Classification MVP", theme=gr.themes.Soft()) as demo:
+with gr.Blocks(title="Company Custom Themes Classification MVP", theme=gr.themes.Soft()) as demo:
     gr.Markdown("""
-    # 🎯 Lucidya Custom Themes Classification - MVP
+    # 🎯 Company Custom Themes Classification - MVP
     
     **A scalable, cost-effective solution for multi-client theme classification**
     
@@ -320,7 +320,7 @@ with gr.Blocks(title="Lucidya Custom Themes Classification MVP", theme=gr.themes
         gr.Markdown("""
         ## 🎯 Solution Overview
         
-        This MVP demonstrates a **hybrid embedding-based approach** for Lucidya's Custom Themes feature:
+        This MVP demonstrates a **hybrid embedding-based approach** for Company's Custom Themes feature:
         
         ### ✅ Key Advantages:
         1. **Cost Effective**: ~$500/month vs $30,000/month for pure LLM approach
@@ -348,7 +348,7 @@ with gr.Blocks(title="Lucidya Custom Themes Classification MVP", theme=gr.themes
         
         ---
         
-        **Built for Lucidya's Case Study | Scalable Theme Classification MVP**
+        **Built for Company's Case Study | Scalable Theme Classification MVP**
         """)
 
 # Launch the app