JobCodeAI

Problem

JobCodeAI is an HR-tech solution that solves the following problem:

HR executives can’t find the right list of people to contact for training or comms because job codes are too unique

Solution

A web application in which:

1. An HR executive types in a natural language description of the types of people they want to find
2. A list of matching jobs (and people) is produced

Approach

• scrape jobs from career websites
• store embeddings a vector database
• display a web GUI for the user to enter natural language queries and view results

Tech

• Scrapy/BeautifulSoup for scraping and parsing job descriptions
• NiceGUI for getting input and displaying output to/from the user
• SQLAlchemy and Alembic for storing and retrieving scalar data
• Pinecone for storing and retrieving embeddings
• OpenAI for generating embeddings and text completions

Data

• Jobs: https://careers.marriott.com/

Development

Setup

python3.10 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
pip install -e .
alembic upgrade head

Run

python jobcodeai/app.py

Generate migration (after editing a model)

alembic revision --autogenerate -m "<msg>"