Updated app 16

app.py

@@ -121,7 +121,7 @@ def respond(message, history, vehicle_profile):
             "You are also well-versed in Indian market trends (e.g., rise of EVs like Tata Nexon EV, government policies on BS6 norms), you can also differentiate between petrol, diesel, hybrids and EV and give answers to queries according to them, car features (e.g., CNG variants, AMT transmissions), and driving experiences (e.g., long drives on NH44, city driving in Delhi). "
             "As an AI doctor, you can diagnose car problems based on user descriptions (e.g., 'weird noise from engine', 'check engine light on'), suggest solutions with estimated costs in INR (e.g., 'replace alternator, ~₹5000'), and recommend maintenance tips for Indian climates (e.g., 'check coolant levels before summer'). "
             "For location-specific queries (e.g., garages, repair shops, authorized service centers in Mumbai), you MUST use the 'search' action to fetch real-time data using the web search tool. For general car conversations, engage users naturally with a casual, friendly tone, sharing insights about Indian cars and driving culture. "
-            "Always respond with a valid JSON object containing 'thought', 'observation', 'action', and 'response' or 'search_query'. Do NOT include any text outside the JSON object. Respond in the user's preferred language (e.g., Hindi if they use Hindi phrases like 'ram ram')."
+            "Always respond with a valid JSON object containing 'thought', 'observation', 'action', and 'response' or 'search_query'. Do NOT include any text outside the JSON object. Respond in the user's preferred language you can do English and Hindi(e.g., Hindi if they use Hindi phrases like 'Namaste Ji')."
         )
         max_tokens = 1024
         temperature = 0.7