Quantify Your Deep Work And Defeat Distractions
Track and improve how you work by using Al-powered software that observes your screen. This tool helps you stay focused and turn scattered bursts of work into more efficient and more productive work sessions.
Designed for privacy, your data never leaves your machine.

MakerTime Demo Reel
Features
Designed for Privacy
By default, no data ever leaves your machine. If you optionally configure OpenAI, your data will only be sent directly to the OpenAI servers, not to any intermediaries.
Local LLM by default
By default MakerTime uses Llama3 running in Ollama. This runs directly on your Apple Silicon CPU, no external service required.
Optional OpenAI backend
You can configure it to use OpenAI as a backend and choose your favorite OpenAI model.
Vision Language Model (VLM)
Optionally leverage a powerful Vision Model (VLM) to analyze your screen activity, such as the one built into ChatGPT4-o or Llama 3.2. See the How It Works page for more details.
OSX native OCR API
Leverages the native OCR API in OSX to extract text from your screen and pass it to the LLM for analysis.
Customizable definition of Deep Work
Customize the LLM prompt template to define what Deep Work looks like for you.
Daily report and goal tracking
At any point you can check the amount of daily deep work you have completed so far. You can also set a goal to trigger a full-screen popup when you reach your daily goal.
Optional Distraction Blockers
You can optionally configure and tune distraction blockers to display a full screen window when you're getting distracted.
Configurable Work Schedules
Set your usual work hours to only analyze your screen during these periods.
Free*
The alpha version of MakerTime is completely free for any commercial or private use. Future versions may have a paid tier, but early adopters will be eligible for deep discounts. Send a message with your email address for instructions to register as an early adopter.
Frequently Asked Questions
How can I verify that no data leaves the machine?
To verify that no unexpected network connections are being made, you can run a firewall like Radio Silence or Lulu.
Is this available as Open Source?
The full source code is not currently available, but the source code for some components is available from an earlier iteration of the project. See tech stack and libraries for details. Reach out if you're interested in contributing to a potential open source release.
Is this project angel or VC backed? What is the revenue model?
Currently a boostrapped part-time indie hacker project. The revenue model is TBD.
What is the power consumption?
When using a local LLM the power draw is unfortunately quite high (numbers TBD). When using the OpenAI API it is much lower. There is a configuration option to only analyze your screen when the computer is plugged in.