JobCheck

Inspiration

Over 14 million people have been scammed by fake job posts within the last year, and with the rise of AI, there has been an increase of 118% fake job posts in 2023 alone. Since applying to a job requires submitting personal information, when people fall for fake job posts, they give away plenty of information they shouldn't otherwise be giving away and could have their identity stolen. We wanted to solve this problem and lower the number of people who are being tricked into giving information about their identity to these fraudulent job posts.

What it does

JobCheck is a google chrome extension that will scrape a linked in job post for its relevant data and then tell the user whether our AI thinks if the job post is fraudulent.

How we built it

Using HTML, CSS, JS, Node, we created a frontend google chrome extension that users can download easily. We used FastAPI to create an API endpoint for the frontend to query if the job post is fraudulent. Then using tensorflow, recurrent neural network (RNN) with long short-term memory (LSTM) layers for text classification, and our two kaggle sets we found (totaling at approximately 28000 job posts) we trained an AI model to detect fraudulent job posts with a 99% accuracy rate.

Challenges we ran into

Finding the technique to create the model was challenging because this was the entire teams first time working with data sets and using data science or ML training to solve problems. Furthermore, this was also our first time developing a chrome extension. Learning all these new techniques and skills was a tough obstacle to overcome.

Accomplishments that we're proud of

We are extremely proud of our end product because as we see it, we think it is good enough to be put on the chrome extension store and be used by users. The learning process and how we were able to figure out everything was another part of this hackathon we were proud of.

What we learned

We learned about using LLMs to assist in ML training, how to build chrome extensions, how to use jupiter notebook, and so many more.

What's next for JobCheck

We are planning on releasing it onto the chrome extension store to be used by others, and we also plan on trying to implement a feature where users can help further fine tune our AI through feedback through the extension.