Align AI

Inspiration

Hours spent sitting, scrolling, and working on screens cause our posture to deteriorate slowly, often without us even noticing. This loss of body awareness means posture problems often go uncorrected until discomfort or pain appears.

Inspired by the idea that technology shouldn’t just correct our behavior but help us understand ourselves, this project reimagines posture tracking as a tool for rebuilding proprioception. By combining AI-powered body analysis with visual feedback and awareness training, the app encourages users to reconnect with their posture, develop better habits, and regain a sense of alignment between their body and the digital world they spend so much time in.

What it does

AlignAI uses AI-powered pose analysis to monitor the user’s posture through a live skeletal overlay, measuring head, spine, and shoulder alignment to detect posture drift in real time. It records daily posture snapshots and heatmaps so users can track how their posture changes over time, while also providing personalized exercises and recommendations based on the specific misalignments detected.

Beyond simple posture correction, the app trains the user’s proprioception by asking them to estimate and adjust their posture without guidance, then compares their guess to an AI-generated ideal posture model to improve body awareness. Over time, the system learns the user’s patterns and aims to predict posture drift before it happens, helping users develop long-term posture habits rather than relying on reactive alerts.

How we built it

We built the interfaces for this project entirely on Figma Make.

Challenges we ran into

One of the biggest challenges was translating complex biomechanical measurements into something simple and understandable for users. Posture can be measured through many different angles and alignments, but showing too many metrics at once quickly becomes overwhelming, so we had to carefully decide which measurements were most meaningful and how to visualize them clearly.

Another challenge was designing a system that trains proprioception rather than just reacting to bad posture, which required creating interactions that encourage users to sense and adjust their posture without relying entirely on visual guidance.

Finally, we had to consider privacy and technical limitations of camera-based tracking, ensuring the system could analyze posture accurately while keeping data secure and designing features that push beyond what current consumer posture apps typically offer.

Accomplishments that we're proud of

We designed an app that use the camera to monitor spinal curvature and posture in real time. We developed a progress tracking feature that allows users to visualize posture improvement over time with body icons and analytics.

What we learned

We learned how AI can be used to analyze human posture and body alignment. We also learned how technology can help improve physical wellness and body awareness.