AlignAI

Inspiration

When it comes to self-directed exercises and trying to improve oneself, it can be difficult to see any progress with your efforts. Stretching and working on flexibility and mobility is one of those things, that I believe is very valuable in today's world. However, without a coach, it is up to our own judgment as to whether or not we are doing the stretches correctly.

That is why I wanted to build AlignAI, a self-directing coach that people can use as a tool for tracking how effective their stretches are, and if they are doing it correctly.

What it does

AlignAI takes inputs from the user about their general soreness across different parts of the body, suggests an exercise to do, and anaylzes a 10 second live recording of the user. It does this through generative AI, which prompts an llm with the data to then generate a score from 0-100, a summary of their performance, and any next-step tips that the user could greatly benefit from.

How we built it

This was built with a react native frontend paired with expo, since I wanted this to be something people could easily install on their phones, so people can spend like 5 minutes a day analyzing their stretches, and then move on to more pressing events.

FYI, more than 70% of the code was AI-generated.

Challenges we ran into

The major challenge was trying to figure out what features to implement. Initially, I had big inspiration to use machine learning to very accurately grade the users stretching position, but with my current technological knowledge and the little time I had, I was forced to pivot to an alternative solution. This led to the idea of using a llm to generate the grades, which proved to be a much simpler and faster solution, that still provides high quality user experience.

Accomplishments that we're proud of

I'm proud of the fact that I built a full scaled app that people could actually use on a day to day basis. While it doesn't have all the features I would want to implement, it is certainly something I am proud of building, especially with very little knowledge of the tech I used.

What we learned

I learned a lot about the tech used, like react native frameworks and components, expo and how I can easily use it on my phone, as well as how to use an llm within my project.

I also gained better time management and decision making skills. In a limited amount of time, I was forced to make decisions to either cut certain aspects of the original idea, or apply alternative solutions that fit within the time I had.

What's next for AlignAI

In the future I would love to implement some of the features I was hoping to do. Most notably, the analyzing of the stretching position through machine learning and data, as well as a higher quality ui that genuinely is appealing to the user.

Built With

• android
• expo.io
• javascript
• node.js
• openai
• openrouter
• react
• typescript
• vercel