GPT-Resume

A basic Django and Nextjs-based Resume filtering application using OpenAI API.

GPT-Resume allows the user to upload resumes (.pdf files) of multiple applicants with the role and job description. It then filters out the suitable candidates for the job. The content of the resume is analysed by GPT-3.5 Turbo model by default. However, you can change the model and it's parameters to tailor your needs.

See the final report: https://github.com/akshanshbhatt/gpt-resume/blob/master/REPORT.md

Setup

Step 0: Pre-requisites

Make sure you have python, git, node, and npm installed locally on your system.

$ python --version
Python 3.12.2
$ git --version
git version 2.43.2
$ node --version
v21.6.2
$ npm --version
10.4.0

You can run these commands to check whether you have them installed or not. The version should be close to mine; not having precisely the same version is not an issue. Just make sure you don't work with very old versions. Make sure you have configured git on your device.

Optional: It would be better if you have conda cli installed on your system for virtual environments and Python version management. This is optional since you can anyway use the native venv module in Python (Although I personally hate it).

Step 1: Cloning the repository

(Preferred Way) If you have ssh keys already set up on your device, then run -

$ git clone git@github.com:akshanshbhatt/gpt-resume.git

on your terminal emulator of choice. Otherwise, clone using https -

$ git clone https://github.com/akshanshbhatt/gpt-resume.git

Step 2: Creating Python virtual environment

After cloning the repository, move into the root directory of the project and run the following commands -

$ cd gpt-resume
$ python -m venv gpt-venv

$ source gpt-venv/bin/activate

If you get no errors, then you have just created and sourced into (activated) a new virtual environment. To make sure, just run -

$ which python
{base_path}/gpt-venv/bin/python

You have done everything correctly if you get the gpt-venv in your path.

(Alternate/Preferred Way) If you have conda cli installed on your system, just run -

$ conda create --name gpt-venv python=3.12

$ conda activate gpt-venv

After creating and sourcing in the virtual environment using any of the two ways, make sure you have the latest version of pip installed -

$ python -m pip install --upgrade pip

Step 3: Installing the backend dependencies

Move inside the gpt_resume directory. This folder contains the backend (server-side) source code. We will be installing the external Python packages required for running the backend using pip.

$ cd gpt_resume
$ python -m pip install -r requirements.txt

In case any dependency causes an issue during the execution of the program, you can install the exact same versions of the dependencies with which I worked -

$ python -m pip install -r requirements.lock.txt

Step 4: Setting up environment variables for the project

You will be required to create a new api key on your OpenAI dashboard if you don't already have one and reference it before starting up the application. DO NOT share this api key with anyone. If you already have an existing api key, you can use that as well. Apart from this, you will also have to generate the api key for your Django application, which is responsible for validating sessions and cookies for the backend. Run -

$ cd ..  # Move to the root of the project before running this
$ python -c 'from django.core.management.utils import get_random_secret_key; print(f"django-insecure-{get_random_secret_key()}")'

to generate a random api key for your Django application. Both of these keys must be referenced in the .env file at the root of the project. Your .env file should look like this -

OAI_API_KEY = "<YOUR_KEY_HERE>"  # Your OpenAI API key. This is only shown once on the dashboard, so make sure you save it. Otherwise, you'll have to generate another one.
DJANGO_SECRET_KEY = "<YOUR_KEY_HERE>"  # This is the random string that you just generated in the terminal. PASTE IT AS IT IS FROM THE TERMINAL.

Step 5: Installing the frontend dependencies

Change your present working directory to gpt_resume_frontend. This folder contains all the frontend (client-side) source code of our application. We'll be using npm cli to install the necessary node modules.

$ cd gpt_resume_frontend
$ npm i

Step 6: Running database migration scripts

Before starting up our backend server, we have to ensure that all the data models are configured with our sqlite database. Move back again into the backend directory and run the migration scripts -

$ cd ../gpt_resume
$ python manage.py makemigrations
$ python manage.py migrate

Step 7: Starting the backend server

To start the backend server, just run -

$ python manage.py runserver

Your server should now be live on port:8000 (http://127.0.0.1:8000/).

Step 8: Starting the frontend application

To start our frontend application, open a new terminal session and move to the gpt_resume_frontend directory (once again). Run -

$ cd gpt_resume_frontend  # Make sure you are inside the frontend directory
$ npm run dev

to host our frontend app locally (in dev setup) on port:3000 (http://127.0.0.1:3000).

That's all. Now go ahead and upload as many resumes as you want for analysis!

Important Note: Make sure that the backend is hosted on port:8000. The frontend will not be able to communicate with the backend if the port is different since these are hardcoded in the frontend code.

Project Structure

Directory	Summary
gpt_resume	Contains all the backend code (Python) of the application.
gpt_resume/gpt_resume	Contains the basic configurations and settings of the Django backend.
gpt_resume/api	The main Django app that handles all the backend functionalities.
gpt_resume_frontend	Contains all the Frontend code (TS/React) of the application.
gpt_resume_frontend/app	Contains the homepage/root route (page.tsx) and layout (layout.tsx) of our fronted.
gpt_resume_frontend/components	Contains all the reusable frontend app components related to the app. Our frontend utilizes shadcn ui components, which is present in gpt_resume_frontend/components/ui.
gpt_resume_frontend/pages	Contains the pages of routes other than the root route. This app uses page-based routing using nextjs.
sample_resume	Contains a few sample resumes for testing the app.

Backend API Endpoints

You can access these endpoints in your browser since the backend uses Django Rest Framework (DRF) views for interactive object creation and visualization. You can also use api testing platforms like Postman or Hoppscotch if you want.

API Endpoint	HTTP Method(s) Allowed	Purpose/Comment
/api/get-applicant-list/<uuid:job_u_id>/	GET	Gives a list of applicants for a particular job posting (referenced by its u_id).
/api/get-applicant-summary/<uuid:u_id>/	GET	Gives the Summary of an applicant (profile, academic experience, professional experience, etc.) after extracting it from the resume. This summary is specific to a particular job posting only.
/api/post-resume/	POST	Allows the user to post multiple resumes (multiple .pdfs as formdata) under files header along with a job_u_id header for generating summary for all the applicants corresponding to a unique job.
/api/get-job-list/	GET	Gives a list of all the jobs (Job objects) posted.
/api/post-job/	POST	Allows the user to post a job using the job_title and job_description headers.
/api/post-resume-with-job/	POST	Combines the functionalities of both /api/post-job/ and /api/post-resume/ under a single endpoint.

External References

Sample Resumes Taken From: https://www.cmu.edu/career/documents/sample-resumes-cover-letters/sample-resumes_scs.pdf