OptiFi - UC Berkeley Hackathon Finalist

OptiFi

At UC Berkeley's AI Hackathon, my team and I created OptiFi, an all-in-one health app that uses AI to reduce the mental friction required to build healthy habits. We built a multitude of features, from computer vision calorie tracking to virtual AI health diagnostics. We were awarded as 1 of 8 finalists for the AI For Good Award and a $25,000 investment out of 1500+ competitors.

My Role

Serving as the product manager for my group, I led the full ideation process for the product, created an interactive high-fidelity prototype in Figma (40+ frames), and pitched our project to 20+ executives and judges (including Caroline Winnett and Sibyl Chen of the SkyDeck team).

Key Features

  • Personalized Habit Tracking

    • Recommends healthy activities to plan in your schedule based on your diagnosis results with GPT-4o and automatically adds accepted events into your calendar with Google Calendar API.

  • Nutrition Scanning

    • Using the camera app on any phone, snap a picture of any food and Anthropic’s Claude 3.5 Sonnet via the Amazon Bedrock API will analyze and log the amount of calories in each meal.

  • AI Medical Diagnostic and Virtual Doctor

    • Utilizing OpenVINO on Intel Corporation's Tiber Developer Cloud, we provide healthy food recommendations based on previous meals.

  • Fitness Metric Tracking

    • Displays all of your important health data in one easily consumable format to track health progression using NumPy and Pandas.

Frames from the deck:

See the full presentation here.

Devpost

More by Nicholas Chua

View profile