Skip to content

Ask any questions about the first book in the series The Golden Compass: Northern Lights

Notifications You must be signed in to change notification settings

francoabaroa/ask

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

79 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Ask My Book

Things You'll Need

  1. OpenAI API Key
  2. Resemble API Key

Setup

  1. Fork and clone repository locally

Backend Setup

  1. cd ./backend
  2. Create and fill in .env using .env.example as an example with the correct API keys
  3. Install dependencies with command bundle install
  4. Run db migrations with command rails db:migrate

Frontend Setup

  1. cd ./frontend
  2. Create and fill in .env.local using .env.sample as an example
  3. NEXT_PUBLIC_API_URL= should be set to your Rails backend url (most likely http://localhost:4000 if you start locally using instructions below)
  4. Install dependencies with command npm install

Please note:

  • If you change frontend port (default is port 3000), make sure you update backend/config/application.rb CORS middleware to accept your new localhost url
  • If you deploy the frontend to Vercel make sure to add it to the CORS middleware config as well
  • If you deploy the backend, make sure you update /frontend/package.json "build" line under scripts object to include your backend deployment URL:
"scripts": {
    "dev": "next dev",
    "build": "NEXT_PUBLIC_API_URL=<YOUR_BACKEND_DEPLOYED_URL> next build",
    "start": "next start",
    "lint": "next lint"
  },

Run Locally

  1. In /frontend directory run npm run dev command to start frontend and will start in port 3000 unless you change it
  2. In /backend directory run rails server -p 4000
  3. Navigate to http://localhost:3000 to access app

Change Embedded PDF

If you want to change the embedded pdf, make sure it's titled book.pdf

Run the following command in the /backend directory

rails runner lib/tasks/format_pdf_manuscript.rb <PATH-TO-book.pdf>

Don't forget to update the book title name in /frontend/pages/index.tsx and relevant book info (author name, bio, header, questions) in backend/app/controllers/api/v1/questions_controller.rb#L122

Potential Improvements

  • Use OpenAI embeddings to find similar questions that have already been asked (currently only checks for exact question in the DB)
  • Add upload PDF functionality on frontend so a user can land on the app, upload their PDF, and ask it any questions (without having to update the code repo)

Testing

  • No tests currently, but when I do implement I will test for the following:

Backend Testing

  • Use Rails' built-in testing framework for writing model, controller, and integration tests
  • Test model validations, associations, and custom methods
  • Test controller actions and API endpoints
  • Test edge cases and error handling

Frontend Testing

  • Use a library like jest for writing tests for React components
  • Test component rendering, user interactions, and API calls
  • Test edge cases and error handling

Architectural Decisions

  • Frontend: NextJS, React, and Tailwind CSS were chosen for frontend development. NextJS provides an easy-to-use framework for building server-rendered React applications, which allows for better SEO and faster initial page loads. React facilitates the creation of reusable UI components and simplifies state management, making it easier to develop and maintain the app. Tailwind CSS is a utility-first CSS framework that enables rapid styling of components and ensures a consistent design throughout the app.
  • Backend: Ruby on Rails was chosen for backend development due to its ease of use, rapid development capabilities, and extensive library support. Rails follows the MVC (Model-View-Controller) pattern, which promotes separation of concerns and makes it easier to manage and scale the application.
  • API: A single API endpoint (POST /api/ask) was created to handle user questions. This endpoint receives the user's question, searches the book's embeddings for the answer, caches the answer, and sends a request to OpenAI's Embedding API. This design simplifies the communication between the frontend and backend, making it easier to maintain and scale the app.
  • CORS configuration: To enable communication between the frontend and backend, CORS (Cross-Origin Resource Sharing) was configured in the backend. This allows the frontend to make requests to the backend API from different origins (e.g., localhost during development and the production domain).
  • Modular components: The frontend components (QuestionInput, AskQuestionButton, and AnswerDisplay) were designed to be modular and reusable. This makes it easier to maintain and update the app, as changes to a specific component won't affect other parts of the application.

Something Broken?

  • Open an issue and I'll get back to you

About

Ask any questions about the first book in the series The Golden Compass: Northern Lights

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published