Inspiration

Traditional productivity apps don't work. They are easy to circumvent, and the stick is too light, as the consequences are nonexistent. As a result, productivity is lowered and we get less done. That's why we created Sisyphus, the world's first AI-native productivity app that both verifies and incentivizes you to finish your tasks.

What it does

Sisyphus is a tongue-in-cheek productivity web app that turns task completion into a “push the boulder” challenge: users first fill out a multi-section onboarding questionnaire, then create a queue of timed tasks and must prove they’re actually working via periodic camera/photo check-ins. The browser client manages the questionnaire flow, task queue, and countdown timers, while a Node/Express backend uses Anthropic’s Claude vision API to analyze submitted images and decide whether the user appears to be working on the stated task. If they pass, the timer advances to the next task; if they fail, the system triggers a consequence by sending an email to a recipient specified in the questionnaire, notifying them that the user did not complete their task on time. The result is a humorous but technically grounded demo of multimodal AI, browser APIs, and server-side automation used to enforce accountability.

How we built it

This project was built on Node and Express.js, and uses the Claude API to verify you're actually working on your tasks by analyzing photos from your webcam or uploaded images.

Challenges we ran into

One of our biggest challenges was synchronizing work among each of the four team members. We were able to safely do this by collaborating on github and assigning separate elements to people to avoid merge conflicts.

Accomplishments that we're proud of

We were able to successfully link email functionality to the website, so that it could "contact your ex" as a punishment, of course, this is all jokes. We are very proud that we were able to put together an app for such an innovative idea, modularizing feature development so that the app can be flexibly developed with the possibility of adding many more features (to be talked about) in the future. During the development of the app, we were so stoked about the potential for this app and about how we would implement this into our daily lives.

What we learned

We learned that claude AI has robust and extensive development applications and that the future of software dev can be accelerated efficiently with its use. In the future, hopefully AI agentic workflows like we've seen today will be integrated deeply within the tech work space to increase accessibility and productivity of coding.

What's next for Sisyphus

We want to expand the range of consequences for users to be specific to their questionare answers and also allow them to specify what level of consequence they want. For better societal impact, we want to implement the feature of donating money to charity for tasks uncompleted.

Built With

Share this project:

Updates