Palette

main page
work in progress - quality of life updates
work in progress - quality of life updates

Inspiration

Palette was inspired by the desire to simplify dining choices by providing a comprehensive overview of restaurants through user reviews and health standards. We wanted to empower diners with not just opinions, but also objective health data, making it easier to choose places that are both loved and safe.

What it does

Palette leverages the power of Yelp reviews and public health data to provide a detailed health score for restaurants. It analyzes emotional content in reviews using Hume's API and computes health scores by combining this data with health inspection records, processed through You.com and OpenAI technologies.

How we built it

We integrated Hume for emotional analysis and utilized You.com with OpenAI for processing health inspection data and scoring. The backend was developed in Python, and we maintained a responsive user interface with React, ensuring a seamless transition between user inputs and our analytical outputs.

Challenges we ran into

We ended up pivoting from our original news aggregation idea, requiring a reevaluation of our data sources and analytical methods. Bridging emotional analysis with objective health data presented unique challenges, especially in aligning the outputs of Hume's emotional analysis with the structured health data processed through You.com and OpenAI.

Accomplishments that we're proud of

We're proud of building this solution to provide tangible difference in upgrading the quality of life for restaurant reviews services. Integrating sophisticated AI tools to create a user-friendly platform that not only informs but also reassures users about their dining choices, has been particularly rewarding.

What we learned

Hume, You.com, OpenAI, AWS, API (Yelp)

What's next for Palette

As Palette evolves, we plan to expand our service to cover more geographical areas and languages. We will refine our algorithms for deeper emotional insights and integrate predictive models to forecast future health scores based on trend analysis and historical data. We also would like to make a more relevant dish recommendations so customers are more prepared as they enter a new restaurant.