Ashley is an AI-powered personal assistant that understands Turkish voice commands and responds appropriately using speech recognition and natural language processing techniques. The project combines Speech-to-Text (STT), NLP, and a simple user interface to create a practical and interactive voice assistant application.
The main goal of this project is to develop a smart assistant that can:
- understand spoken Turkish commands,
- analyze the user’s intent,
- execute the related action,
- and return the result both visually and audibly.
- 🎙️ Turkish Speech Recognition (Speech-to-Text)
- 🧠 Intent detection using rule-based NLP
- 📝 Note creation and note listing
- ⏰ Reminder management
- ➗ Basic mathematical calculations
- 🔊 Text-to-Speech responses
- 🖥️ Streamlit-based user interface
- 📦 Modular and extensible project structure
| Area | Technology |
|---|---|
| Programming Language | Python |
| Speech-to-Text | Vosk (Turkish model) |
| Natural Language Processing | Rule-based NLP |
| Text-to-Speech | pyttsx3 |
| User Interface | Streamlit |
| Data Storage | JSON |
| Supporting Libraries | NumPy, SciPy, SoundDevice |
ℹ️ Note: The Whisper library was not used due to dependency incompatibilities with Python 3.13. Instead, Vosk was preferred as a stable, offline, and Windows-friendly speech recognition solution.
Ashley/
├─ app.py
├─ README.md
├─ requirements.txt
├─ data/
│ ├─ commands.json
│ ├─ notes.json
│ └─ reminders.json
├─ models/
│ └─ vosk-model-small-tr-0.3/
├─ modules/
│ ├─ speech_to_text.py
│ ├─ text_to_speech.py
│ └─ intent_handler.py