iRecuit

Inspiration

What it does

[A recruiting website strictly for students who fulfill diversity requirements, where users, in this case, students (and recruiters) have profiles, and students (users) get connected to the latest internship opportunities, based on their categories of interest. Every candidate application is done in the Application room of that job and candidates can get more listings based on the most clicked title feature. Candidates "may" be able to track the progress of each application to see which is Accepted, Rejected or Pending.]

How I built it

I used postgreSQL for the database on Express server. I had two other servers. One for processing texts before going into the ML model and one for caching.

Challenges I ran into

• Could not use NLP, natural and other Python based ML libraries
• Built ML algorithms with JavaScript
• Performance optimization using session memory
• Bad data from API for TF-IDF calculations

Accomplishments that I'm proud of

Despite the limitations I still was able to pull it off and timely

What I learned

Pros and cons of TF-IDF, Naive Bayes, KNN and approaches like content based filtering and collaborative filtering

What's next for iRecuit

Building the recruiter side and integrating it into the database, such that they are the admins and can edit the outcomes of applicants’ statuses and maybe a chat platform between applicants and recruiters

Built With

• css
• html5
• javascript
• jest
• react
• vite