Cure-ify

Cure-ify is data-driven personalized food and recipes recommendation progressive web app enhanced with soothing herbs and spices food items and recipes.

Comment

Inspiration

Herbs and Spices have been long standing home remedies to cure sickness not only complimentary treatments for the symptoms of chronic condions like menopause: arthritis, bone density loss, hot flashes, and anxiety.

So food of interest with nutritional content can not only act as optimal nutrition for recovery but for preventative medicine. We chose to focus on chronic illnesses because other conditions are, by nature, random, thus producing too sparse of a signal to track.

A patient often compromises with the food items which are tasteless and something with does not align with their interest. Food being so importantly co-related with the mood of the person and hence with their physiological processes.

What it does

Cure-ify intakes data from user in the form of questionnaire which are as follows: Q0. What type of physiological help you seek from the food? (In form of health target options) a. Digestive b. General Health c. Brain d. Infection e. Cardiovascular f. Female Health

Q1. Do you have any allergies ? Q2. What are your dietary restrictions? Q3. How are you feeling now?

Q4. What would you like to eat? (Answer could be abstract as much possible like, in speech)

   eg - "Something **chicken**, **soup** and **asian**"

From the above sentence the AssemblyAI will transfer the speech to text and the highlighted words will be taken as keywords and the recommendation system will search through the datasets with respect to the given keyword.

The datasets have the (food, recipes, time to make, calories, no. of ingredients, all ingredients). From this dataset we will recommend user a food choice. The answers would be taken as JSON inputs which will be passed onto the the backend through the API for the processing.

From the input user data we will recommend food - guessing the user interest from the speech of input.

The user will have the food, and on three days we will get a feedback from the user about the effects of that food on their body/physiological system).

Each time user will receive food, the feedback system will be due in 36 hours. Feedback will be in form of a rating scale from (0 - 10). Upon user feedback we will track their personalized food interest, impact of the recommended food on their physiology, and track the progress and update the type of recommendation prediction. Hence making the sick-meal planning personalized and beneficial for the user.

How we built it

1. Frontend

Frontend is built with progressive react.js which is compatible in running on every device like PC, iPad, mobile. We use node.js for scalability and it can support multiple concurrent requests and it an excellent choice for real time data streaming applications. We used CSS 3 for the makeup construction for the webpage. We used Virtual Studios as IDE.

2. Backend

Backend is written in Python on Jupiter Notebook with multiple APIs usage with AssemblyAI, MATLAB for Data Viz. We used excel to make and manage datasets. We used Firebase of Google Cloud to manage data and employ processing on the data.

*3. Datasets And Databases *

We used excel to create and get datasets. We used FIREBASE of Google Cloud to manage the data and its utilization to develop web app , get data analytics and maintenance purpose.

Databases:

  1. Toast : - Its good for notifications
  2. ReactToastify : - Can get your MP3 clips.
  3. AXIOS : - It can make http requests.

4. APIs

We used AssemblyAPI for speech to text recognition. We used API to generate datasets on spreadsheet - from the user feedback loop.

5. Technology

Matlab :- For data vizualization AssemblyAI :- For speech to text recognition

*6. Google cloud Domain names *

    cureify.ai

Challenges we ran into

Teammates confirmations

We had a teammate confirmed for the hackathons but after the opening ceremony - they abruptly said they are working with someone else so won't be working with us.

We had similar issues - after one night of coding and developing newly 400 lines of code, teammate decides to just run out of the hackathon and without committing to the GitHub. So we lost 8 hours of work, instantly.

AssemblyAPI - API token 401 error

When we tried to integrate the Assembly AI into our system - the API token generated from the Assembly AI -caused us 401 response error by the mentor had left the hacking arena.

Google Cloud Credits Problem

GCP didn't accept the school email and constantly kept asking for credit card information which was unavailable for some time and had to go via the personal email route to access the FIREBASE.

Accomplishments that we're proud of

Project accomplishment

We had hit a moment where we thought completing the project wouldn't be possible. Out team was disintegrated and demotivated to work. We recomposed ourselves and gained the spirit and developed the application inspire of all challenges.

Organizing the team in the face of defeat

We had lost complete hope but we decided to act as team re-structured the architecture and distributed the task among the team members.

Sleepless but still motivated to complete the task at hand

Our team members were religious follower of healthy sleeping habits from 10pm to 6am. So they really fought super hard to stay awake and work with small naps on table and inbetween work breaks.

Brainstorming : From Vague to Inspirational

In the early stage - we didn't like the idea of the food health. But in spirit of hackHarvard theme of "Take something from past and make something out of it" we used my project from TreeHacks '22 - FoodMood and our teammates project from hackRice '22 SpiceMD and took inspiration from them to develop a personalized sick meal planning app.

What we learned

  1. Never depend on one person for the tech stack they say - they would handle, rather have back up.
  2. Get something very basis started real quick which can be developed in interation.
  3. Discuss efficiently to make demo product.
  4. Fast critical decision making skills is highly necessary in the real world dev.
  5. Play to your strength.

What's next for Cure-ify

  1. Continue in R&D about the Big-data formation and database management for personalization records of the individual user.
  2. Making B2C business model running on google ads.
  3. Develop mobile android app and iOS app.

Built With

Share this project:

Updates