• License
• Installation
• Setup
• Usage
• Discussion
• Credits
• Contact
• Acknowledgements
• Project Status
• Project Team

• NO LICENSE

1. Create an empty directory for the project.

mkdir CSai

2. CD into the directory.

cd CSai

3. Create a virtual environment in the directory (don't forget to activate the VE).

python3 -m venv venv

4. Clone the project repo

git clone https://github.com/BigMoonTech/CSai.git

5. Install the requirements

pip install -r requirements-dev.txt

1. You will need to rename config-sample.py to config.py and add your own Openai API key.
2. In the same config.py file, you will need to add your own mail server/gmail credentials.
   1. If you are using gmail, it used to be that you had to enable less secure apps. This has been discontinued. Instead, you will have to enable 2-factor authentication and generate an app password. You can find more information here: https://support.google.com/accounts/answer/185833?hl=en

1. Run the app (port 5000)

flask --app src --debug run

2. Optionally, you can run the app like so (port 5006):

python3 app.py

3. Open the app in your browser by pasting this (if port 5000):

http://127.0.0.1:5000

Note: If you are running the app on port 5006, you will need to change the port number in the URL. Also, you should see something like * Running on http://127.0.0.1:5000 inside the record.log file that will be generated in your main project directory.

• This project was built as a final project for CS50, Harvard's Introduction to Computer Science course. It is a full-featured application that allows users to create an account, verify their account through email, login, update their personal information, and interact with the application.
• CSai utilizes Openai's API to generate responses to user input. The user can ask questions about computer science and the application will respond with a response generated by a GPT-3 model that is fine tuned to act like a computer science tutor.
• The project was written mostly in Python, and built using the Flask framework. It uses SqlAlchemy to interact with a database that stores user information, and information about OpenAi completions.
• The project uses Bootstrap 5 and custom CSS for styling, and Jquery and Javascript for some of the front-end functionality.

• CSai (or whatever name is given to the main project folder which holds everything)
  • src (source code folder)
    • db
      • database file
    • db_models
      • database models and db_session file. db_session is used to interact with the database and uses sqlalchemy to create the database, tables, and rows.
    • helpers
      • helper functions
    • infrastructure
      • cookie_auth
        • cookie authentication
      • request_dict
        • A module that allows you to access request.form, request.args, request.headers, and route args as a dictionary
      • tokenizer.py
        • Serializes and deserializes data to send url safe strings to verify accounts
      • view_modifier.py
        • Defines a decorator that is used to modify the view function. Instead of rendering templates everywhere in view methods, I return a dictionary with the request variables etc. to the Jinja2 templates. I use this pattern throughout the project. Basically, when decorated with @response, the view method returns a dictionary, the decorator takes that dictionary, and creates a response with the template specified, and the dictionary values. That is what is rendered from a view method.
    • services
      • user_service.py
        • Handles user logic. This includes creating a user, verifying a user, and updating user info, such as, how many free calls they have left. It also handles, password hashing, and password verification.
      • completion_service.py
        • Handles OpenAi completions. Searching, creating, adding completions to the db etc.
    • static
      • all the static files
    • templates
      • all the HTML templates
    • view_models
      • The view models are defined here. The view models are used to pass data to the templates in a more organized way, and used within a view method. After a view method's logic plays out, I turn all the model's attributes into a dictionary and return the model from the view method, which then goes to the @response decorator, as stated above.
    • views
      • The views are where logic is controlled. Inside a view method, I create an instance of one of the view_models, perform some logic depending on whether it is a post or get request, and then return the view model's attributes as a dict.
  • init.py
    • Where the app is configured.
  • config.py
    • This is where the environment variables and other configurations are defined
  • app.py
    • Used to run the application with Python3
  • tests
    • all the tests for testing the app are here
    • conftest.py is used in pytest to share a pytest fixture accross the tests.
    • test_client.py is used to define a pytest fixture, in this case a fixture primarily used to config the app for testing

1. training.talkpython.fm
2. realpython.com
3. stackoverflow.com
4. openai.com
5. flask.palletsprojects.com
6. getbootstrap.com
7. jquery.com
8. colorhunt.co
9. google fonts
10. sqlalchmey.org
11. python.org

• Email: [email protected]
• Github: BigMoonTech
• LinkedIn: https://www.linkedin.com/in/joshua-aguilar-8677b532/

• training.talkpython.fm was a very useful resource for learning how to use and write some of the infrastructure that the app is centered on. Particullary the request_dict, cookie_auth, and view_modifier modules.
• realpython.com was a great resource for learning how to implement email verification in a flask app.

• In Development
• In Production
• In Maintenance

• Joshua Aguilar