Project Info

AI Posture Guardian

Devpost

This project did not submit a demo video on Devpost.

Inspiration

AI Posture Guardian was inspired by the way laptop work slowly pulls people into bad posture without them noticing. We wanted to build a productivity tool that does more than remind you politely in the background. It actively protects your focus and your body by stepping in when your posture slips. What It Does AI Posture Guardian uses the laptop camera to detect posture in real time. When it sees sustained slouching or uneven shoulders, it shows a fullscreen reminder that blocks the screen until the user sits up straight again. The blocker can also be dismissed by pressing the spacebar twice as a manual override. How We Built It We built the project as an Electron desktop app with a React frontend. MediaPipe Pose tracks body landmarks from the laptop camera, including shoulders and ears. A posture engine evaluates head position, shoulder alignment, and shoulder tilt to decide whether the user is sitting upright. Electron powers the fullscreen always-on-top overlay, while React handles the live dashboard, sensitivity controls, warning timer, and camera preview. We also disabled Electron background throttling so posture detection keeps running even when the user is working in another app. Challenges We Faced The hardest part was tuning the posture detection so it felt helpful instead of annoying. If the thresholds were too strict, the app blocked the screen constantly. If they were too loose, it missed real slouching. We iterated on shoulder-angle detection, recovery timing, and manual unblock behavior to make the experience feel more natural. Another challenge was making the blocker work while the app was in the background. The app needed to keep detecting posture even when the user was focused on another browser or tool, so we had to adjust Electron behavior to keep the camera loop active. What We Learned We learned how to combine real-time computer vision with desktop app behavior in a way that feels immediate and practical. We also learned that UX matters a lot for wellness tools: a posture app should be firm enough to change behavior, but forgiving enough that people actually want to keep using it. What's Next Next, we would add daily posture analytics, calibration for different body types and camera angles, voice coaching, and privacy-first local session history. We could also add integrations for productivity workflows, like pausing distracting sites when posture drops or generating weekly posture improvement summaries.

Analysis

Compare with all teams

View

Metric

Figures cover GitHub contributors during the hackathon window. A co-authored commit counts in full for each author, so per-member totals add up to more than the whole-team figures.

Technology

Found in codeClaimed only
  • CSSIn code
  • HTMLIn code
  • JavaScriptIn code
  • ReactIn code
  • Node.jsClaimed

4 of 5 appear in the indexed code. 1 claimed on Devpost could not be matched to code, which may simply mean the tool leaves no trace in the repository.

AI coding agents

No AI coding agent signals were found in this repository.

Detected from committed agent config files and commit authorship. Absence of a signal is not proof an agent was unused.

Codebase size

Source size

19 KB

Source files

8

Counts recognized source files only; vendored directories, binaries and lockfiles are excluded, so this is smaller than the repository on disk.

0 stars