Socially Distanced Pantry

COVID-19 has affected communities around the world in ways we haven't experienced before and our at-risk and low income individuals are in most need of help.

We created a crowd-sourced utility to provide information about the virus around you.

With current data, we hope you can make the best decisions.

You can help others by offering the extra supplies you might have and letting people know where to pick it up then the people that need it can search for the supply they most need and where to find it.

An informative and highly scalable web application covering Covid-19 and encouraging the sharing resources with the community that is available without an internet connection the instant a user loads the app.

Quick start

Make sure that you have Node.js v8.15.1 and npm v5 or above installed.
Clone this repo using git clone --depth=1 https://github.com/brandonjurado/Covid19Hack.git <YOUR_PROJECT_NAME>
Move to the appropriate directory: cd <YOUR_PROJECT_NAME>.
Run npm install in order to install dependencies.
At this point you can run npm start to see the example app at http://localhost:3000.

The Stats/ML Models Frame Work with Login Server Load Balancing

In the interest of quick delivery of information with almost infinite scalability, we decided to employ the "Kubeless" archictecture so that login systems would never become bogged down, and we could then provide a fast and easy mechansim for data scientists and others to get information out quickly and efficiently. This is built on top of AWS Lambdas and Serverless.

To build we are going to assume that you have a kubernetes cluster already set up and configured with kubectl and eksctl. You will also need AWS Lambdas setup with AWS Fargate and the CLI tools installed and ready to go. From there, you are going to run:

$ kubectl create ns kubeless
$ kubectl create -f https://github.com/kubeless/kubeless/releases/download/v1.0.6/kubeless-v1.0.6.yaml

This will take some time, but it creates a new namespace called kubeless and then creates services in that namespace. Once completed install:

$ git clone https://github.com/serverless/serverless-kubeless
$ npm install serverless -g

This will get the plugin from serverless for kubeless and install it globably. After that, your Kubernetes and AWS administrators will need to provide you with IAM user profiles as well as config files that need to be put in the proper place. Once everything is done, we have provided an example of the boiler plate endpoint setups, and when you have modidifed them to your needs simply run: $ serverless deploy It will automatically take care of provision of everything from S3 to lambdas and provide you with a reverse proxy to the endpoint!

Contributors

_{Armando Soriano}

_{Brandon Jurado}

_{Michael Osei}

_{Preston Ward}

This project follows the all-contributors specification. Contributions of any kind welcome!

License

Built With

Submitted to

EarthxHack 2020

Created by

Worked on idea creation and AWS integration

Armando Soriano
An enthusiastic programmer with a passion for learning.
I worked on setting up the back end and producing the machine learning/statistical models. While thinking of what architecture to design for the back-end, I remembered seeing a graph that showed the incredible amount of page visits and higher than normal stresses on banking systems when people were logging in en masse to check if their stimulus had come in. This inspired me to create an "infinitely scalable" load balancer for logins utitizling AWS lambdas to simply act as a function that would authenticate the user with the server, alleviating a lot of stress for the system. I then deployed a Kubernetes cluster with the Kubeless plugin to handle the rapid creation and deployment of end points that could be designed with arbitrary functions. The hope is that State Farm or any business could wrap this deployment mechanism so that data scientist and other specialists could quickly and efficiently deploy statistical and ML models to all of their users keeping them better informed quicker, and without any downtime. Finally, to help realize the last goal, I used serverless with the serverless-kubeless plugin to bring the developyment and deployment process down to just two commands, where you only needed to define your .yaml file to serve your model, as seen on our apps main page. The bottom picture is a model produced that shows what would happen if we release social distancing to early, and the top is a general aggregate stastic at the county level of rates of COVID19.

psward Ward
Worked on adding scalability/availability for offline use of application during peak load times.

Brandon Jurado
Just an ambitious, self-driven developer making contributions to open-source and as a full-time developer.

Updates

Armando Soriano started this project — Apr 26, 2020 04:55 PM EDT

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.