iPromise - AI focus buddy for deep work.

Product Introduction

1. Definition: iPromise is a macOS-native desktop application (category: Productivity/Utility) that functions as an AI-powered focus assistant and accountability partner. It is a deep work protocol built on the psychological principle of Body Doubling, using on-screen AI characters and real-time screen context analysis to monitor and enhance user focus.
2. Core Value Proposition: iPromise exists to bridge the gap between task planning and task execution. Its primary value is in providing real-time accountability and intervention to prevent digital distraction during deep work sessions, thereby increasing actual productivity and reducing procrastination for knowledge workers, students, and remote professionals.

Main Features

1. AI Focus Buddy with Screen Context Analysis: The core feature is an animated AI companion (characters include a goose, turtle, rabbit, comet, or tree) that resides in the MacBook's notch or floats on-screen. It uses local screen recording to analyze the active window's context in real-time. This analysis, powered by a user-provided AI API key (e.g., OpenAI, Anthropic), detects shifts from a focused task to distractions (e.g., social media, news sites). When a distraction is detected, the buddy provides a gentle, contextual nudge to bring the user back on track. How it works: Screenshots are processed locally and sent directly to the user's chosen AI provider for analysis; iPromise's servers never see or store this visual data.
2. Structured Focus Sessions with Post-Session Analytics: Users initiate a session by making a "promise"—selecting a single task (draggable from other apps) and setting a timer. During the session, a Heads-Up Display (HUD) shows the focus state. Upon completion, the app generates an AI-powered session summary and evaluation. The Session Inspector provides a detailed 60-segment focus bar, a timeline with segment summaries, distraction insights, and suggestions for improvement, enabling quantitative review of focus patterns.
3. Privacy-First Architecture with iCloud Sync: iPromise employs a unique privacy model where all sensitive screen analysis data is processed locally or through the user's own AI API. No screenshots or screen recordings are uploaded to iPromise's servers. Non-sensitive data—such as focus history, session summaries, streaks, and insights—is synchronized across the user's Mac devices via iCloud sync, ensuring a seamless experience without compromising privacy.

Problems Solved

1. Pain Point: The "planning-execution gap" where individuals can plan tasks in to-do apps but consistently fail to execute due to digital distractions and a lack of immediate accountability during the work session. This leads to context switching, reduced deep work time, and procrastination.
2. Target Audience: Primary users include software developers, writers, researchers, students, and remote knowledge workers who require sustained, uninterrupted focus. It also appeals to individuals familiar with productivity methodologies like Pomodoro or Body Doubling but seek a more automated, interactive, and tech-integrated solution.
3. Use Cases: 1. Writing a Technical Report: A researcher uses iPromise to block out 90-minute sessions for writing, with the AI buddy intervening if they switch to academic social media or email. 2. Coding a Complex Feature: A developer starts a session for a specific module; the buddy helps maintain flow state by discouraging context switches to Slack or Hacker News. 3. Studying for Exams: A student commits to 50-minute study blocks, using ambient sounds and the buddy's presence to simulate a focused study hall environment.

Unique Advantages

1. Differentiation: Unlike passive timer apps (e.g., Pomodoro timers) or restrictive website blockers, iPromise is proactive and contextual. It doesn't just block sites; it understands on-screen content and intervenes intelligently. Unlike traditional to-do apps (Todoist, Things), it is not for planning but for dedicated, guarded execution of a single planned item.
2. Key Innovation: The integration of local, privacy-preserving screen context analysis with a user-configurable AI provider. This allows for highly personalized and intelligent distraction detection without the privacy risks of sending screen data to a third-party service. The AI Buddy as a "Body Double" is a software innovation that gamifies accountability and reduces the isolation of solo deep work.

Frequently Asked Questions (FAQ)

1. Does iPromise record my screen or store my screenshots? No, iPromise uses macOS screen recording permissions solely to capture context for real-time, local analysis. Screenshots are processed locally and sent only to your own configured AI provider (e.g., OpenAI API). iPromise's servers never receive, see, or store any of your screen data.
2. What AI models does iPromise support, and do I need my own API key? iPromise is designed to work with your preferred AI provider. You must supply your own API key from a supported provider (such as OpenAI's GPT-4 or Anthropic's Claude). This model gives you control over costs, privacy, and the specific AI's capabilities used for your context analysis and session summaries.
3. Can I use iPromise on my iPhone or iPad? Currently, iPromise is a macOS-only application available for Mac. The product description indicates compatibility requires macOS 14.0 or later, and its core functionality relies on deep system integration with the Mac desktop environment and notch display.
4. How does the AI Focus Buddy actually detect distractions? The app periodically captures the content of your active window. This data is sent to the AI model you configured via your API key. The AI is prompted to analyze whether the on-screen content aligns with your stated task promise or constitutes an unrelated distraction, triggering the buddy's nudge mechanism based on the response.
5. What happens to my focus data and history? Your session history, focus streaks, and AI-generated text summaries are stored locally on your Mac and synchronized across your devices using iCloud. This metadata allows you to track progress and review insights. Crucially, this synced data does not include any of the original screen captures or visual data used during the sessions.