Obatin - Self-Medication

Obatin is a machine learning-based application that is able to provide advice or how to use drugs that are freely available at drugstores and drugs that can be purchased without a doctor's prescription. Obatin can be used by teenagers to adults who already have awareness and responsibility in the use of over-the-counter drugs.

Obatin Teams - C22-PS234

Name	Path	Bangkit ID	Social Media
Azzahra Novitri Wulandari	Android	A2159F1688
Ridian Putra	Android	A2159F1687	Instagram & Github
Raihan Nur Ramadhan	Machine Learning	M7010F1098	LinkedIn & Github
Mohamad Rayhan Adhari Nugroho	Machine Learning	M7010F1089	LinkedIn & Github
Ragil Murdiantoro Aji	Cloud Computing	C2159F1689	LinkedIn & Github
Eko Cahya Purnama	Cloud Computing	C2152G1670	LinkedIn & Github

Screenshots

Splash Screen	Opening Page	Login Page

Register Page	Home/Chat Page	Profile

Infrastructure Design

Android Documentation

• GitHub Repository

• Example Apps

  UI/UX

  • Figma Design

Machine Learning Model Installation Locally

These are the steps in installing a machine learning model locally

Install Requirements

• Install Flask Framework
• Install Matplotlib
• Install Numpy
• Install Sastrawi
• Install Tensorflow

1. Run Locally

git clone https://github.com/raihannur45/obatin.git

• How To Run Code

python simulasi.py

curl http://localhost:5000/bot?text=cara%20minum%20obat%20bisolvon

2. Run using docker container images

• Install Docker
• Pull the Container Image

docker pull ekocahya/api-ml:v1
docker run -d -p 5000:5000 ekocahya/api-ml:v1

curl http://localhost:5000/bot?text=cara%20minum%20obat%20bisolvon

Installing Obatin-API Server Locally

These are the steps in installing a machine learning model locally

Install Requirements

• Install Node.JS and NPM
• Install Memcached
• Install PM2
• Install PostgreSQL

1. Run Locally:

• Create a new database with a name that you can customize which will later be used to migrate the configuration of the existing tables. For the type of database itself using SQL, you can use other applications for the database of the same type.
• Clone This Repo

git clone https://github.com/Ian-72/Obatin-API.git

• Go to Obatin-API directory

cd Obatin-API

• Install the required modules

npm install

• Make sure to set the .env file if you want to run this code in Production and if you want to run in development set the ./src/config/config.js file
• Run migrate

npm run migrate

• Start API server
  • Run In development

    npm run start-dev

  • Run In Production

    npm run start-prod

• The default port set in the Obatin-API server for development is 5000 and for production it is set in the .env file. Make sure the running ports for the machine learning model and API server are different.

2. Run using docker container images

• Install Docker
• Pull the Container Image

docker pull c2159f1689/c22-ps234-cc-api

• Run The container

docker run -d --name cc-api -e NODE_ENV= -e MIGRATE= -e HOST= -e PORT= -e DB_HOST= -e DB_USERNAME= -e DB_PASSWORD= -e DB_DATABASE= -e DB_DIALECT= -e MEMCACHIER_SERVERS= -e ACCESS_TOKEN_AGE= -e ACCESS_TOKEN_KEY= -e REFRESH_TOKEN_KEY= -e ML_API= -p 8080:8080 c2159f1689/c22-ps234-cc-api

• Example

docker run -d --name cc-api -e NODE_ENV=production -e MIGRATE=0 -e HOST=0.0.0.0 -e PORT=8080 -e DB_HOST=34.121.94.31 -e DB_USERNAME=postgres -e DB_PASSWORD=postgres -e DB_DATABASE=obatindb -e DB_DIALECT=postgres -e MEMCACHIER_SERVERS=172.25.0.3:11211 -e ACCESS_TOKEN_AGE=900 -e ACCESS_TOKEN_KEY=cecaf4a11f439bf4699cd36c68de0792a6bc741729cfad5a93f1d367f59257f69c153adb128a359b727dba3ef928ea245ee0ff19e5be44342e9aa9d5df1cf1c7 -e REFRESH_TOKEN_KEY=cb389a2d32bb567baf3a792d5e6253ccea80bb95185f3418fad645001f09bb8cfe295e63a072d8b28daa14e59c317fd69c6ead61275a634791b233e73ca1a51f -e ML_API=http://172.17.0.2:5000 -p 8080:8080 c2159f1689/c22-ps234-cc-api

How to Deploy the code to Google Cloud

Requirements:

• Configure the VPC

  All instances that will be created will be on the same network in the same zone, you need to configure the VPC if you want to use the network you created yourself, if you can't use the default network that already exists. For more details about VPC, see Documentation at this link.

• Cloud SQL for PostgreSQL

  Creating a PostgreSQL instance in Google Cloud, make sure to choose the appropriate zone to use to store the instance, or you can visit the Documentation link provided by Google for instance creation.

• Memorystore - Memcached

  Creating an instance of Memcached on Google Cloud, make sure to select the same zone as the other instance to use to store the instance, or you can visit Documentation link provided by Google for instance creation.

• Compute Engine

  Creating a virtual machine in Google Cloud, make sure to select the same zone as the other instance to use to store the instance, or you can visit Documentation link provided by Google for instance creation.

1. Configure PostgreSQL Instance:

In this section you have to configure which networks the instance can accept, go to the CLoud SQL page and click the created instance name, then go to the Connections page and search for "Authorized networks", then enter CIDR from your virtual machine's IP.

2. Configure VM's:

At this stage you can configure by imitating the steps in the local installation listed above. And make sure to use the IP and port on the Memcached instance and PostgreSQL instance for use in configuring .env files or Docker commands.

3. Set up Firewall Rules:

So that the application can be accessed on the internet, we must open the port for the virtual machine by setting it in the firewall on the VPC network that we are using. For an explanation and how to configure the firewall, you can visit this Documentation.

Obatin-API Reference

Authentications

endpoint /auth

Endpoint	Method	Parameter	Type	Description
/auth	POST	application/json	string	This endpoint used for user login. Payload contains username and password field, you can fill of username with username or email, after you fill payload and send request to server, server will response 201 and data contains accessToken and refreshToken.
/auth	PUT	application/json	string	This endpoint used for update access token, example when accessToken has expired. Payload contain refreshToken field, after you fill payload and send request to server, server will response 200 and data contains new accessToken.
/auth	DELETE	application/json	string	This endpoint used when user logout. Payload contains refreshToken fields, after you fill payload and send request to server, server will response 200 with status success.

Bot

endpoint /bot

Endpoint	Method	Parameter	Type	Description
/bot	POST	application/json	string	This is endpoint where you can talk with Obatin BOT. Payload only contains text field, after you fill text field and send request to server, if success, server will response 200 and data contains response from BOT, else server will response 400 if you send bad payload or server will response 401 that mean you need to login first
/bot/message/	POST	application/json	string	This endpoint is used when the label of the model is known but the name of the drug is not known or has not been entered by the user.
/bot/message/reply	POST	application/json	string	This endpoint is used when the label of the model is known and the drug name is known and will return the answer according to the requested drug name.

Users

endpoint /users

Endpoint	Method	Parameter	Type	Description
/users	POST	application/json	string	This endpoint used for user register. Payload contains email, password, passwordConfirmation, username, and fullname field, after you fill payload and send request to server, server will response 201 and data contains user id.
/users/profiles	GET	application/json	string	This endpoint is used to get user profile information based on user name or user email. You must attach email or username in the query parameter, after you fill it and send request to server, server will response 200 and data contains user profiles.