Multimodal · AI2024
AI Assistant
Voice-first AI assistant combining GPT-4 for chat and DALL·E 3 for image generation with speech I/O.
- Role
- Full-stack engineer
- Timeline
- 2024
- Stack
- FlutterOpenAISpeech-to-TextTTS
01
Screens & Architecture
02
The Problem
Voice-first AI assistant combining GPT-4 for chat and DALL·E 3 for image generation with speech I/O.
03
How I built it
- Voice UI with STT and TTS for hands-free use.
- GPT-4 chat orchestration and DALL·E 3 image generation.
- Prompt scaffolding for consistent tone and length.
04
Outcome
- 95% recognition accuracy on target commands.
- Portable playground for multimodal experiments.