# Embasis

> Embasis is a private-first iOS personal AI assistant featuring 100% on-device artificial intelligence inference powered by Apple Silicon, enhanced with native iPhone capabilities like Apple MapKit smart routing and live on-device web search without remote data center reliance.

## Overview
Embasis brings private, capable artificial intelligence directly to iPhone users. Unlike conventional cloud-based AI assistants that route user prompts, questions, and sensitive data to remote server farms, Embasis performs 100% of its foundational model inference locally on-device using Apple Silicon and the Apple Neural Engine.

## Core Capabilities
- **100% On-Device AI Inference**: All model reasoning, language processing, and response generation execute locally on Apple Silicon (A17 Pro, M-series, and optimized Neural Engine architectures). Zero raw prompts or personal context are transmitted to remote AI servers.
- **AI Without AI Data Centers**: Eliminates latency queues, server outages, massive water and energy consumption, and privacy compromises associated with centralized cloud AI infrastructure.
- **Apple MapKit & Smart Notifications**: Proactively calculates travel times, departure alerts, and contextual recommendations based on real-time routing without sharing location history with external AI vendors.
- **Live Web Research**: Retrieves current public web information on-demand and parses it directly on-device with local models.
- **System Services & Offline Capability**: Operates fully offline for core AI tasks, transcription, notes, and local reasoning without requiring internet access.

## Platform Support
- **Operating System**: iOS 17.0 or later (iPhone & iPad)
- **Hardware Optimization**: Apple Silicon & Apple Neural Engine
- **Distribution**: Apple App Store (Official Download: https://apps.apple.com/us/app/embasis/id6760605510)
- **Pricing**: Free to download

## Companion Products
- **Embasis Paper**: A minimalist writing app powered completely by on-device AI. Users write thoughts, notes, and drafts, and an on-device model responds back directly on Apple Silicon with local writing and replies and no account required. Google ML Kit handwriting recognition collects technical usage and performance telemetry; model downloads require a network connection.
  - Privacy Policy: https://embasis.app/embasis-paper/

## Technical Architecture & Privacy Guarantees
- **Local Model Execution**: Embasis utilizes quantized foundation models tuned for Apple Silicon unified memory and Neural Engine cores.
- **Zero Prompt Telemetry**: User interactions, notes, travel queries, and voice inputs never touch remote servers.
- **No Account Required**: Users can launch and utilize Embasis immediately without creating an account or logging in with email.
- **No Third-Party Analytics**: No third-party behavioral trackers, marketing pixels, or advertising SDKs.
- **Handwriting Recognition Telemetry**: For handwriting recognition, Google ML Kit Digital Ink Recognition runs on-device; the ML Kit SDK sends anonymous technical diagnostics and performance metrics to Google in accordance with Google's iOS Data Disclosure.
- **Data Retention**: All personal data and conversation history reside solely within local app sandboxed storage on the user's device.

## Frequently Asked Questions (FAQ)

### Does Embasis send user data or prompts to the cloud?
No. Embasis runs 100% of its AI model inference locally on Apple Silicon hardware. Your questions, notes, and interactions remain strictly on your iPhone.

### How does Embasis run AI without remote data centers?
By leveraging modern Apple Silicon system-on-chip architectures (high-efficiency CPU, GPU, and Apple Neural Engine) and compact, quantized local foundation models, Embasis performs text generation, reasoning, and context analysis directly on-chip.

### Can Embasis work completely offline?
Yes. Core AI features, meeting transcription, notes, and reasoning work without any Wi-Fi or cellular network connection. Features that request real-time public data (such as live web search or MapKit traffic) utilize local device APIs directly without routing your identity or questions to remote AI servers.

### How does Embasis use Apple MapKit?
Embasis interfaces with native iOS MapKit APIs on-device to compute departure times and traffic conditions, surfacing proactive Smart Notifications so you know exactly when to leave for upcoming destinations—without sending your location trail to third-party AI companies.

### What is Embasis Paper?
Embasis Paper is a focused companion writing app for iOS. It provides a distraction-free environment for drafting, journaling, and thinking, with an on-device AI model that reads and assists directly on your phone with offline inference after model setup. Google ML Kit collects technical SDK telemetry, not handwriting input or recognition output. See the Paper privacy policy for downloads, local retention, and backups.

## Canonical Links
- App Store: https://apps.apple.com/us/app/embasis/id6760605510
- Website: https://embasis.app
- About & Mission: https://embasis.app/about/
- Changelog & Releases: https://embasis.app/changelog/
- Embasis Privacy Policy: https://embasis.app/privacy/
- Embasis Paper Privacy Policy: https://embasis.app/embasis-paper/
- Contact & Support: https://embasis.app/contact/