Deploy SupportPulse demo: Gradio triage interface with pre-computed results

app.py

@@ -178,7 +178,7 @@ def run_triage(subject: str, body: str) -> tuple:
   </div>
 
   <div style="font-size: 11px; color: #475569; text-align: right;">
-    SupportPulse AI Pipeline | {datetime.now().strftime('%H:%M:%S')} | Model: gemma2:2b + LightGBM SLA
+    SupportPulse AI Pipeline | {datetime.now().strftime('%H:%M:%S')} | Model: gemma4:e4b + LightGBM SLA
   </div>
 </div>
 """
@@ -248,7 +248,7 @@ def run_triage(subject: str, body: str) -> tuple:
   <h3 style="margin: 0 0 12px; color: #f1f5f9;">AI-Generated Grounded Response (RAG)</h3>
   <p style="line-height: 1.7; color: #cbd5e1;">{rag_response}</p>
   <div style="margin-top: 12px; font-size: 12px; color: #475569;">
-    Generated by gemma2:2b grounded on {len(result.get('similar_tickets',[]))} retrieved historical tickets
+    Generated by gemma4:e4b grounded on {len(result.get('similar_tickets',[]))} retrieved historical tickets
   </div>
 </div>
 """
@@ -282,7 +282,7 @@ with gr.Blocks(title="SupportPulse Intelligence Platform") as demo:
         SupportPulse Intelligence Platform
       </h1>
       <p style="color: #94a3b8; margin: 10px 0 0; font-size: 1.1rem;">
-        End-to-End MLOps | LLM Cascade &middot; RAG &middot; ChromaDB &middot; LightGBM &middot; FastAPI
+        End-to-End MLOps | LLM Cascade (gemma4:e4b) &middot; RAG &middot; ChromaDB &middot; LightGBM &middot; FastAPI
       </p>
       <div style="margin-top: 16px; display: flex; gap: 10px; justify-content: center; flex-wrap: wrap;">
         <span style="background: #1e3a5f; color: #60a5fa; padding: 4px 12px; border-radius: 20px; font-size: 13px;">68,235 tickets indexed</span>