Inspiration
In today's era, prioritizing health is paramount. Self-improvement remains central.
What it does
- Nutritional Analysis: Provides comprehensive calorific data for specified food items, alongside curated recipes featuring the input item.
- Custom Workout Plans: Tailor's workout regimens based on the user's age, desired workout frequency, and current weight.
- Dietary Planning: Recommends meticulously crafted meal plans for both vegetarian and non-vegetarian diets, aligning with the workout plan.
- Recipe Compilation: Construct detailed recipes for suggested meals, accompanied by precise calorific breakdowns.
How we built it
Leveraging Amazon Bedrock Playground, I have crafted a Fitness and Wellness application. Through adept utilization of Amazon Bedrock's array of user input widgets, we engineered a streamlined input mechanism. Employing the CLAUDE model, we synthesized system outputs, with select outputs interfacing with diverse LLM model widgets.
Challenges we ran into
The overarching challenge lies in refining the app's specificity. To address this, we harnessed user-centric parameters such as age, weight, and workout frequency. However, further iterations are imperative to fortify reliability.
Accomplishments that we're proud of
As a novice in machine learning, my successful creation of an app utilizing LLMs stands as a testament to my dedication. The app holds promise in aiding fitness enthusiasts in embarking on structured fitness journeys, complemented by dietary discipline.
What we learned
The hackathon afforded invaluable lessons in prompt engineering and the meticulous construction of fitness-oriented machine-learning models. This experience lays a robust foundation for future endeavours in consumer-facing product development.
What's next for Fittter
Looking ahead, our focus shifts towards refining the app beyond generic offerings. The objective is to furnish personalized fitness and wellness solutions accessible to all demographics. This necessitates augmenting the app with additional features tailored to individual user requisites.
Built With
- amazon-bedrock-platform
Log in or sign up for Devpost to join the conversation.