Startup Idea Generator

This web application generates ideas for your next billion dollar startup company. Ideas are generated in the form "<idea> for <thing>".

Why?

This is a toy web application and server that demonstrates (most of) the methodology of the twelve-factor app. It features:

• A simple Python Flask server for the API and Angular app for the frontend.
• Easy integration with MongoDB, but trivially extendible to other databases.
• Dev and production parity for easy development and deployment.
• Almost no custom CSS by leveaging Bootstrap and Angular Material.
• Amazon Elastic Beanstalk configurations for deployment.
• Github actions for full CI/CD.

Topology

This application is distributed into three parts:

1. Frontend
   • Angular application that is main point of interaction for the user.
2. Admin console
   • Separate Angular application to interact with the database.
3. API
   • REST API written in Python using Flask.
   • Enforces user access control for admin API.

Here is what the service and request topology looks like in a production environment using AWS and MongoDB Atlas:

In local development, the environment is setup to work without DNS routing snd MongoDB Atlas is replaced with Docker running mongodb.

Development Getting Started

For development you will need to install:

• Docker
• Python (>=3.9)
• Node.js
• angular/cli

Start the mongodb instance with docker. You can do this in the infrastructure directory:

> docker-compose up

This is just for development. Do not use in production!

Use a tool like Robo3T to check the mongodb instance is working. This will also be useful for inspecting data later.

Alternatively, you could use MongoDB Atlas for your development database.

Next, install the StartupGeneratorAPI package to a virtualenv with --editable. Before we start the API application, we will set up some data in our mongodb instance. Create the following python script (or run it in a python console):

import pymongo
from StartupGeneratorAPI.utility import add_words_to_db_from_text

client = pymongo.MongoClient('mongodb://words-writer:dev@localhost:27017/app')
add_words_to_db_from_text(r'.\tests\data\word_set_1.txt', 0, client)
add_words_to_db_from_text(r'.\tests\data\word_set_2.txt', 1, client)

Consider using an IDE like PyCharm for this.

After running the above, use Robo3T (or similar) to check documents were added to the words collection.

Let's set up the environment variables the API application needs. Use your platform's syntax for setting environment variables:

export admin_username=root
export admin_password=root
export mongo_reader_uri=mongodb://words-reader:dev@localhost:27017/app
export mongo_writer_uri=mongodb://words-writer:dev@localhost:27017/app
export flask_secret_key=not_safe_for_deployment_change_me
export serve_static_from=redirect

Change these for production!

Start the API:

> startup-generator-api api

Go to http://localhost:5000 and you will see it's live by rendering:

Startup Idea Generator API

Next step is to serve the admin console at http://localhost:4200. Run this in the admin directory:

ng serve --proxy-config src/proxy.conf.json

Instead of going to that url, go to http://localhost:5000/login where you will see the admin login:

Log in with the username and password set above (root and root), and you will be redirected to the admin console.

And now for the main event! Start the frontend app in the frontend directory:

ng serve --proxy-config src/proxy.conf.json --port 4201

Go to http://localhost:4201 and you will see:

Deployment and CI/CD

Unfortunately we won't cover all that here--that would require several more tutorials! However, checkout the .elasticbeanstalk directories in frontend and api for example Elastic Beanstalk configurations. In an EB production environment, you can set variables from a local machine like this:

> eb setenv mongo_reader_uri=<db_uri>/app

Did you know username and password can be encoded right into the mongo_uri?

The .github/workflows show how everything can be built with GitHub's CI and deployed right to both EB and S3.

Search and replace instances of your-url-here in this repo to set up the configurations for a production environment.

Extra

To increase performance of the db queries, you can run the following as an extension to the above python code:

db = client['app']
db['words'].create_index([
    ('word', pymongo.ASCENDING),
    ('set_index', pymongo.ASCENDING)
], unique=True)