Prior to As Advice, self-help programs either gave inaccurate advice because of inherent differences between people and decisions cannot be made with a fixed algorithm, or it gave inefficient and expensive advice because someone had to manually review your situation and goals. We decided to reform this.
What it does
Our program creates helpful and pragmatic advice given community-driven datasets and short questionnaires.
How we built it
We leveraged the React/Next.JS and Firebase stack to create a front end and back end which interfaces with a scikit-learn based algorithm, which can connect answers to questions with personalized advice.
Challenges we ran into
Figuring out how to efficiently gather data in a way that works both for the algorithm and the end-user.
Accomplishments that we're proud of
- Tree generation
- Our algorithm's ability to perform efficiently and accurately even given a relatively minimal datset ## What we learned
Angelhacks was a learning curve for many of our team members, whether this was learning how to code or efficiently working in a team. But overall, despite setbacks, it was a positive experience and we'd all do it again.
What's next for As Advice
We hope the technology we created in the last couple of days can be improved both by us and by the community and hopefully help improve the lives of many.