A futuristic iOS app that transforms your voice into intelligent AI conversations. Built entirely with SwiftUI and a custom implementation of the TCA (The Composable Architecture) pattern for a clean, reactive, and testable structure. The app leverages Apple’s Speech framework for on-device transcription and integrates with local LLMs via Ollama, enabling private, offline-friendly AI chat.
Featuring a glassmorphic, AI-inspired interface with smooth animations, glowing chat bubbles, and real-time recording feedback, VOICECHAT AI brings a conversational AI experience to your fingertips — powered by your own voice.
-
🎙️ Voice Recognition:
Convert speech to text in real time using Apple’s Speech framework. -
🤖 Local AI Chat:
Communicate with an LLM (like Llama or Phi) through Ollama’s local inference server — no cloud APIs required. -
🧠 Composable Architecture (TCA):
A fully custom, dependency-free TCA setup for predictable state management and clear separation of concerns. -
💬 Futuristic Chat UI:
Elegant, glassmorphic chat bubbles with dynamic gradients and smooth scrolling animations. -
⚡ Private & Offline-Friendly:
All speech processing and AI inference occur locally — your data stays on your device.
VOICECHAT AI follows the TCA (The Composable Architecture) pattern with a clean separation between state, actions, and environment dependencies.
-
View Layer (SwiftUI)
Handles UI rendering, animations, and reactive bindings to app state. -
Reducer Layer
Defines all user interactions, state transitions, and side effects like speech recognition and AI calls. -
Environment Layer
Manages dependencies including the speech recognizer and Ollama network client.
- Xcode 16+
- iOS 17.0+
- Swift 5.9+
- Ollama (running locally)
Download and install from ollama.com, then start the server:ollama serve