On Day 7, we’ll personalize chatbot responses by allowing the avatar to remember user details like their name, mood, or preferences. This creates a more human-like conversation where the assistant feels familiar and attentive.
1. Why Personalize AI Chatbot Responses?
✅ User Engagement: Users enjoy conversations that feel personal.
✅ Memory Retention: Remembering details like names, previous chats, or preferences builds trust.
✅ Contextual Conversations: Allows the assistant to adjust tone and suggestions based on past interactions.
We’ll use: 🔹 AsyncStorage – Store user preferences locally.
🔹 GPT Contextual Memory – Pass chat history + user data to GPT API.
🔹 Emotion History – Track user moods over time.
2. Installing Local Storage for User Data
Step 1: Install AsyncStorage
npm install @react-native-async-storage/async-storage
3. Saving & Retrieving User Preferences
Step 1: Create UserMemory.js Helper
import AsyncStorage from '@react-native-async-storage/async-storage';
export const saveUserName = async (name) => {
await AsyncStorage.setItem('userName', name);
};
export const getUserName = async () => {
return await AsyncStorage.getItem('userName');
};
export const saveMood = async (mood) => {
await AsyncStorage.setItem('lastMood', mood);
};
export const getMood = async () => {
return await AsyncStorage.getItem('lastMood');
};
Step 2: Use Memory in ChatBot Component
Modify ChatBot.js:
import { saveUserName, getUserName, saveMood, getMood } from '../helpers/UserMemory';
const [userName, setUserName] = useState('');
const [lastMood, setLastMood] = useState('');
useEffect(() => {
const loadMemory = async () => {
const storedName = await getUserName();
const storedMood = await getMood();
if (storedName) setUserName(storedName);
if (storedMood) setLastMood(storedMood);
};
loadMemory();
}, []);
Step 3: Save User Name (Once Asked)
Modify sendMessage:
if (input.toLowerCase().includes('my name is')) {
const extractedName = input.split('my name is ')[1].split(' ')[0];
await saveUserName(extractedName);
setUserName(extractedName);
}
Step 4: Save Mood Based on Emotions
If you have voice emotion detection integrated from Day 4:
useEffect(() => {
if (userEmotion) {
saveMood(userEmotion);
setLastMood(userEmotion);
}
}, [userEmotion]);
4. Sending Personalized Context to GPT API
Modify sendMessage:
const sendMessage = async (input) => {
const context = [
userName ? `The user's name is ${userName}.` : '',
lastMood ? `The user's mood was recently ${lastMood}.` : '',
].filter(Boolean).join(' ');
const messages = [{ role: 'user', content: `${context} ${input}` }];
const response = await axios.post(
'https://api.openai.com/v1/chat/completions',
{ model: 'gpt-4', messages },
{ headers: { Authorization: `Bearer ${OPENAI_API_KEY}` } }
);
const botReply = response.data.choices[0].message.content;
setMessages([...messages, { text: botReply, sender: 'bot' }]);
};
Example:
- User: “My name is Sarah.” → Stored as
Sarah. - User: “How’s the weather?” → Sent as: The user’s name is Sarah. How’s the weather?
5. Personalizing the Avatar’s Greeting
Modify AvatarRenderer.js to show a personal greeting:
<Text style={styles.greeting}>
{userName ? `Hi, ${userName}! Ready to chat?` : 'Hello! What can I do for you today?'}
</Text>
6. Testing User Personalization
Step 1: Start the App
expo start
Step 2: Simulate a Chat
- User: “My name is Alex.”
- Assistant (later): “Good to see you again, Alex!”
Step 3: Close & Reopen App
- Avatar should greet the user by their name from local memory.
7. Expanding User Memory
✅ Store Favorite Topics – User likes talking about football.
✅ Track Frequent Emotions – User is often stressed; offer calming responses.
✅ Remember Task History – Remind users of unfinished tasks.
Example:
await AsyncStorage.setItem('favoriteTopic', 'football');
const favoriteTopic = await AsyncStorage.getItem('favoriteTopic');
8. Optimizing Memory Handling
✅ Limit Storage Size – Don’t store large conversations; only keep preferences.
✅ Encrypt Sensitive Data – Use expo-secure-store for storing personal info.
expo install expo-secure-store
9. Key Concepts Covered
✅ Added user memory storage using AsyncStorage.
✅ Personalized GPT chatbot context based on user name and mood.
✅ Improved AI avatar’s greeting & response personalization.
10. Next Steps: Voice Customization & Assistant Personalities
Tomorrow, we’ll:
🔹 Change the chatbot’s voice tone based on mood or personality.
🔹 Let users choose a friendly, professional, or playful assistant.
11. References & Learning Resources
12. SEO Keywords:
Personalized AI chatbot, GPT chatbot with memory, React Native AI assistant, storing user preferences React Native, voice assistant user memory.
