Neural Mail 🦀✨

A high-performance, privacy-first email client built with Rust and Slint, featuring deep local AI integration.

Key Features

• Outlook-Inspired UI: A modern, responsive design with support for Light and Dark modes, resizable sidebars, and fluid navigation.
• Tejas AI Assistant: Chat with your entire 1000+ email inbox using local LLMs. Ask about action items, summarize themes, or find specific invoices without your data ever leaving your machine.
• Reply w/ AI ✨: Generate professional, context-aware email drafts instantly based on the active thread.
• Instant Greetings: Blazing-fast, time-aware conversational responses for simple greetings, bypassing the LLM for better responsiveness.
• Advanced Composition: Support for CC/BCC fields, local attachment management, and integrated email format validation.
• Privacy First: All emails are stored in a local SQLite database, and all AI processing is handled locally via Ollama.

Technology Stack

• Core: Rust
• UI Framework: Slint
• Database: SQLite (via rusqlite)
• Local AI: Ollama (Llama 3.1)
• Networking: Reqwest, Tokio

Getting Started

Prerequisites

1. Ollama: Install Ollama and download the Llama 3.1 model:

   ollama run llama3.1

2. Rust: Ensure you have the latest Rust toolchain installed.

Installation & Run

1. Clone the repository:

   git clone <repo-url>
   cd email-client

2. Run the application:

   cargo run --release

Local Development

The project is structured into three main components:

• src/main.rs: Application logic, Slint callbacks, and event loop.
• src/ai.rs: Local LLM integration handling 32k context windows for deep email analysis.
• ui/app.slint: High-performance UI definitions and layout logic.

---