Support

Q: What is Tahutu Assistant?

A: Tahutu is a voice-based AI assistant with English learning features. It uses on-device speech processing for privacy, letting you practice conversations with AI, build vocabulary with spaced repetition, and save notes for later review.

Q: Is the app free?

A: The app itself is free. However, you need API keys from AI providers which may have their own pricing.

Q: What languages are supported?

A: Currently, the app supports Vietnamese for speech recognition (ASR) and text-to-speech (TTS). When using cloud ASR, multilingual support including Vietnamese and English code-switching is available.

Q: How do I get API keys?

You can obtain API keys from the respective AI service provider's website. The app settings will guide you on which providers are supported.

Q: Are my API keys secure?

A: Yes. API keys are stored using platform-native encrypted storage (iOS Keychain / Android Keystore). They are never sent to our servers.

Q: Can I use the app without API keys?

A: Yes! You have two options:

• Proxy Mode: Sign in with Apple or Google to use our proxy service - no API keys needed
• Direct Mode: Provide your own API keys for direct access to AI providers

Speech recognition and text-to-speech work on-device without any API keys.

Q: What are practice scenarios?

A: Practice scenarios are pre-built conversation exercises to help you practice English in realistic situations. Topics include ordering food, making appointments, job interviews, and more.

Q: How does vocabulary practice work?

A: The app uses a Spaced Repetition System (SRS) to help you remember vocabulary. Words you're learning will be reviewed at optimal intervals, with mastery levels from 1 to 5.

Q: How do I save notes?

A: Tap the notes icon during a conversation to save words, phrases, sentences, grammar points, cultural insights, or pronunciation tips. Notes are organized by category and stored locally on your device.

Q: Is my learning data backed up?

A: Learning data is stored locally on your device. We do not back up this data to the cloud. If you uninstall the app, your learning history will be deleted.

Q: What's the difference between Push-to-Talk and Hands-Free mode?

• Push-to-Talk (PTT): Hold the microphone button while speaking. Recording stops when you release.
• Hands-Free (HF): Tap once to start listening. The app uses voice activity detection (VAD) to automatically detect when you stop speaking.

Q: Why does hands-free mode sometimes cut off my speech?

A: The voice activity detection has a timeout for silence. If you pause too long, it may think you've finished. Try speaking more continuously, or use push-to-talk mode for longer pauses.

Q: Why isn't speech recognition working?

Check the following:

1. Microphone permission is granted
2. You're in a quiet environment
3. Speaking clearly in Vietnamese
4. If using cloud ASR, ensure you have a valid API key and internet connection

Q: Which speech recognition engine should I use?

• On-device: Best for privacy, works offline, Vietnamese only
• Cloud: Better accuracy, multilingual, requires internet and API key

Q: Why are AI responses slow?

Response time depends on:

• Your internet connection speed
• The AI provider's server load
• The complexity of your question

Try using Gemini 2.5 Flash for faster responses, Claude 3 Haiku for balanced performance, or DeepSeek for free tier usage.

Q: The AI gave incorrect information. What should I do?

A: AI responses may not always be accurate. Do not rely on them for critical decisions. Always verify important information from reliable sources.

Q: Is my voice recorded?

A: No. Voice audio is processed in real-time and not stored. When using on-device processing, your audio never leaves your device.

Q: Is my conversation saved?

A: Yes, conversations are saved locally on your device so you can resume them later. You can delete individual conversations or clear all history from within the app. Conversations are never sent to or stored on our servers.

Q: What data is sent to external servers?

A: Only your transcribed text (not audio) is sent to the AI provider you selected. If using cloud speech recognition, your audio is also sent to the respective provider.

Q: The app crashed. What should I do?

Try the following:

1. Close and reopen the app
2. Clear the conversation history
3. Check that your API keys are valid
4. Restart your device
5. Reinstall the app

Q: I'm getting "API Error" messages.

Common causes:

• Invalid or expired API key
• API usage quota exceeded
• No internet connection
• API service outage

Check your API key settings and try again.

Q: Text-to-speech isn't working.

A: The TTS models are loaded on first use. Wait for initialization to complete. If issues persist, try restarting the app.