Discover Dayflow, an open-source macOS app that uses AI to automatically record and analyze your screen activity, generating clear daily timelines with privacy-first design.

Ever reached the end of a day, stared at a neatly arranged calendar full of meetings, and yet completely struggled to recall where your time actually went? Traditional calendars and to-do apps record our plans, not our actual execution. But the screen doesn't lie—and Dayflow, a macOS app, finally makes it talk.

As a macOS user, I've been on the hunt for a tool that could automatically track my time—until I discovered Dayflow, an open-source application that uses AI technology to automatically analyze screen activity and generate daily timelines. Let me walk you through this remarkable tool that's redefining how we log our time.

🎯 What Is Dayflow? More Than a Time Tracker

Dayflow is a native macOS application built with Swift. Its core philosophy is simple yet powerful: If your screen knows what you've been doing, why not let it tell you automatically?

Unlike the manual time-tracking tools we're familiar with, Dayflow fully automates screen recording and combines AI technology for intelligent analysis, generating clear and concise timeline reports. Best of all, as an open-source project (MIT licensed), you retain full control over your data with no privacy concerns.

✨ Features: Smart, Automatic, Privacy-First

🚀 Automatic Timeline Generation

Dayflow records your screen at an extremely low 1 frame per second and uses AI to analyze the footage every 15 minutes, producing concise activity summaries. This means you get a full-day activity timeline without lifting a finger.

🔒 Privacy-First Design

Any app that monitors your screen all day must put privacy and transparency first. Dayflow excels here:

  • 📦 Open-Source Transparency: Full source code is public, so every data-handling process can be audited
  • 💾 Local Storage: All data stays on your device; recordings older than 3 days are automatically deleted
  • 🔧 Flexible AI Modes: Supports both cloud-based Gemini API and fully local processing

⚡ Lightweight and Efficient

When running in the background, Dayflow consumes only about 100MB of RAM and less than 1% CPU, so the vast majority of users won't notice any performance impact. The app itself is just 25MB—remarkably light.

📈 Intuitive Visual Timeline

Once analysis is complete, Dayflow presents your day as a visual timeline, giving you an at-a-glance view of how your time was distributed.

🔬 Core Technology: Balancing Two AI Processing Modes

Dayflow supports two analysis modes to suit different privacy and performance needs:

🌐 Gemini Cloud Pipeline

Leveraging Google's native video understanding capabilities, this mode analyzes video content directly with extreme efficiency—requiring only 2 LLM calls per analysis.

💻 Local Pipeline

This mode reconstructs understanding by analyzing descriptions of individual frames. All data stays on your device, but processing complexity is significantly higher, requiring 33+ LLM calls.

These two modes let users choose based on their privacy requirements and device capabilities—no need to painfully choose between functionality and privacy.

📊 Dayflow vs. Traditional Time Management Tools

🗓️ Compared with Traditional Calendar Apps

Traditional calendars record your plans; Dayflow records your actual execution. This gap is often the most missing piece in time management.

👨‍💻 Compared with Manual Tracking Tools

Tools like ManicTime or Life Cycle can also log time, but Dayflow's AI analysis lets it understand content—not just duration. It doesn't just tell you that you spent 2 hours in a browser; it figures out whether you were writing a report or scrolling social media.

🆚 Compared with Other Screen Recorders

Dayflow's unique advantages include:

  • 🧠 Smart Summaries: Not just recording—AI-generated intelligent summaries
  • 🗑️ Auto Cleanup: No storage worries; recordings are deleted after 3 days
  • 🔍 Open-Source Transparency: Unlike closed-source commercial apps, you can fully trust its data handling

💼 Real-World Use Cases: Who Needs Dayflow?

🎯 Freelancers and Remote Workers

Accurately track actual time spent across different projects, providing data for billing and efficiency optimization.

🎓 Students and Researchers

Understand your study patterns and peak productivity hours to better plan your schedule.

👨‍🎨 Knowledge Workers and Creators

Break through the confusion of "feeling busy all day yet accomplishing nothing" by objectively seeing how your time was allocated.

📱 Self-Quantification Enthusiasts

Add objective time-use records to your personal data analysis without the burden of manual logging.

🛠️ Tips for Getting the Most Out of Dayflow

⚙️ Choosing the Right AI Mode

  • 🔒 Privacy First: Choose local mode (Ollama or LM Studio) to keep data entirely on-device
  • 🎯 Maximum Accuracy: Use Gemini mode for more precise analysis
  • ⚖️ A Balanced Choice: Enable Gemini billing (even on the free tier) to enjoy the data protection of a paid service

📝 Understanding the Data Analysis

Dayflow's analysis isn't simple app-usage statistics—it's true semantic analysis that understands screen content through AI. This means it can distinguish between work-related reading and entertainment browsing.

🔄 Combine with Time Management Methods

Pair Dayflow's data with traditional techniques like the Pomodoro Method or time-blocking. The plan-vs-actual comparison helps you continuously refine how you allocate time.

🚀 Future Outlook: From Recording Tool to Intelligent Management Assistant

Based on user feedback, Dayflow is evolving from a tool that simply records time and tasks into an intelligent assistant that actively helps users stick to their plans. Possible upcoming features include:

  • 🔔 Plan-vs-Actual Drift Alerts: Notifications when you deviate from your planned schedule
  • 📋 Customizable Dashboards: Ability to ask questions about specific work patterns, like "How much time did I spend on code reviews today?"
  • 📖 Daily Journal Feature: Guided reflection prompts to help you consciously review your workday

These additions will transform Dayflow from a passive recorder into a proactive management assistant that genuinely boosts personal productivity.

📥 Installation and Setup Guide

💻 System Requirements

  • macOS 13 or later
  • Xcode 15+ (for building from source)

👥 Installation Steps for Regular Users

  • Download the latest Dayflow.dmg file from GitHub Releases
  • Drag Dayflow into the Applications folder
  • Launch the app. When prompted for permissions, go to: System Settings → Privacy & Security → Screen & System Audio Recording and enable Dayflow

🔧 Building from Source (for Developers)

git clone https://github.com/JerryZLiu/Dayflow.git
cd Dayflow
open Dayflow.xcodeproj

Run the Dayflow scheme in Xcode. If you're using Gemini mode, remember to add your GEMINI_API_KEY under the scheme's "Arguments > Environment Variables."

🛡️ Privacy Setup Recommendations

If you choose Gemini mode and care about privacy, consider:

  • Enabling project billing in Google AI Studio (even on the free tier)
  • This makes your data subject to paid service terms, and Google won't use your prompts and responses to improve its products or models

🔄 Complete Reset

To wipe all data: quit Dayflow and delete the ~/Library/Application Support/Dayflow/ folder.

💫 Conclusion: Rediscover Your Work Patterns

Dayflow is more than a tool—it's a new way of working with awareness. By letting AI quietly observe and learn our digital habits, we gain unprecedented self-insight—not through manual logging, but through intelligent observation.

In an era where attention is the most valuable resource, understanding how we actually distribute our time matters more than ever. Dayflow sits at the intersection of privacy, transparency, and practicality, offering us a digital mirror that doesn't lie.

Your screen has witnessed your work. Now it can finally speak.

🎉 Try Dayflow and uncover the truth about your time.

Dayflow is an open-source project. Feel free to contribute code or submit suggestions in the GitHub repository. This article is based on the project README and hands-on testing; for technical details, please refer to the official documentation.

Note: This is the English translation of the original Chinese version.