Autogig

This is an application that makes it easy for job seekers to find and land interviews at their dream job.

It currently matches candidates with the best jobs based on their preferences (location, experience, and interests) and generates tips for job candidates to bolster their resume.

Room for improvement:

• Fine tune the job matching algorithm to include more parameters
• Handle resume improvements within the PDF itself
• Auto-apply with AI

Getting Started

To get started, clone the repository and install the dependencies:

git clone https://github.com/cypherprotocol/autogig.git
cd autogig
pnpm install

Then, run the development server:

pnpm dev

Supabase

1. Create a Supabase account
2. Create a new project
3. Go to Settings > API
4. Copy the API URL and API Key
5. Follow the Langchain setup instructions to create a vector store

Scraping jobs

1. Create a SERPAPI account
2. Copy the API Key
3. Modify the allJobs.ts script.
4. Run pnpm run all-jobs to scrape jobs
5. Run pnpm run embed to embed jobs into the vector store

scripts Folder

The scripts folder contains additional scripts for manual job scraping using SERPAPI and other APIs.

• jobs.ts script fetches direct job links from provided job_id
• allJobs.ts script fetches all jobs from SerpAPI with a provided query string and location query
• embed.ts script embeds all jobs into the Supabase database using vector embeddings
• deduplicate.ts removes all high similarity jobs from the database, ensuring that only unique jobs are stored (needs improvement)

Stack

The stack being used in the codebase includes:

• OpenAI: Used for generating tips and completing chat prompts.
• Next.js: The framework used for building the web application.
• Tailwind CSS: A utility-first CSS framework used for styling the application.
• Supabase: A backend-as-a-service platform used for creating a vector store for job embeddings.
• Langchain: A library used for generating embeddings and working with vector stores.