AGI, Inc. is an AI lab bringing superintelligence to the edge, on a mission to make it 100% secure, private, and accessible. The company is enabling the era of AI-native devices with agents that are fully private, trustworthy, and run locally — ushering in true decentralization of intelligence. Their first product, AGI-0, is a personalized, proactive AI co-worker that gets things done on your smartphone, now available in early access. The AGI mobile app is available in early preview. Capabilities shown include Taxi, Travel, Message, Music, Order, Shopping, and Delivery.
AGI, Inc. develops the AGI mobile app (early preview) for iOS, featuring AGI-0, a personalized, proactive AI co-worker that runs locally on the smartphone. The iOS engineering team builds and maintains features for on-device AI execution (LLMs, agents running locally without cloud dependency), personalized agent behavior (learns from user's habits, preferences, and context), proactive task execution (initiates actions like ordering taxis, booking travel, sending messages, playing music, ordering food, shopping, coordinating deliveries), natural language interaction (voice or text interface to instruct the AI agent), local data storage (all personal data stays on device — privacy-first), offline functionality (agents work without internet connection), integration with device capabilities (calendar, contacts, messages, music, maps, payment methods, app intents), secure multi-modal input (voice, text, camera for context), custom AI model optimization (Core ML for running LLMs and agent models locally with low latency), and accessibility features (VoiceOver, voice control for agent interaction).
Mobile development in this context focuses on on-device AI execution (Core ML for LLM inference, custom neural networks, transformer models running entirely on iPhone/iPad), model optimization (compression, quantization, pruning for edge deployment with limited memory and compute), local agent orchestration (state machines, decision trees, or learned policies for proactive action selection), privacy-preserving personalization (user embedding stored and processed only on device), real-time audio processing (voice commands with Speech framework, custom wake-word detection), offline functionality (all AI capabilities available without internet — no cloud fallback), integration with iOS system APIs (MapKit for taxis/directions, MessageUI for sending messages, MusicKit for Apple Music control, StoreKit or Payment APIs for orders, shopping, delivery, Calendar for scheduling, Contacts for personalized recommendations), SwiftUI for agent conversation interface (natural language chat or command UI), Push Notifications for proactive agent suggestions ("Want me to order your usual coffee?"), WidgetKit for home screen agent shortcuts, Siri Intents (custom intents for agent commands), App Intents (iOS 16+ for shortcut automation), Core Data for local storage of user preferences, conversation history, and learned patterns, Security and privacy (all data on-device, no cloud upload, encryption at rest), Accessibility (VoiceOver for agent interaction, voice control for users with motor disabilities), and performance optimization for battery-efficient AI (model inference with minimal power draw, background processing).
An AI lab focused on edge superintelligence and decentralized AI like AGI, Inc. hires:
These roles focus on building the AGI mobile app (early preview) that brings superintelligence to the edge with AGI-0, a personalized, proactive AI co-worker that runs entirely on the smartphone — 100% private, secure, and accessible without cloud dependency. Engineers work on on-device LLM inference using Core ML and Apple Neural Engine (ANE) optimization, enabling AI agents to run locally with low latency and low battery drain, agent orchestration (state machines, decision policies) for proactive task execution (taxi booking, travel planning, messaging, music playback, ordering food, shopping, delivery), privacy-preserving personalization where all user data and learned patterns stay on the device (no cloud upload, encryption at rest), offline functionality (AI works without internet — true edge AI), voice interaction with Speech framework and custom wake-word detection, natural language chat interface with SwiftUI, system integration with iOS APIs (MapKit for directions, MessageUI for sending messages, MusicKit for Apple Music control, payment APIs for orders, Calendar for scheduling, Contacts for personalized recommendations), deep linking and app intents for third-party app interaction (Uber, DoorDash, Instacart, Spotify, etc. with user permission), push notifications for proactive agent suggestions ("I noticed you usually order coffee now — want me to do that?"), Siri Intents and App Intents for voice-activated agent commands, Core Data for local storage of conversation history, user preferences, and learned patterns, security and privacy (on-device encryption, secure enclave for sensitive data, zero-trust architecture), accessibility (VoiceOver for visually impaired users, voice control for motor disabilities), and performance optimization for battery-efficient AI (model quantization, pruning, compression, Apple Neural Engine acceleration, memory management) — all while enabling true decentralization of intelligence where AI agents are fully private, trustworthy, and run locally on everyday smartphones, ushering in the era of AI-native devices. AGI's first product, AGI-0, is now available in early access, backed by investors and partners powering the vision to bring AGI to everyday life with capabilities spanning taxi, travel, messaging, music, ordering, shopping, and delivery.
