The problem ATSBot solves

It is a 'Recruitment Assisting' platform which will help recruiters filter out resumes for a particular job profile. Currently, there are hundreds of millions of candidate profiles and CVs online. Manual screening of resumes is still the most time-consuming part of recruiting. 75% to 88% of the resumes received for a role are unqualified. Screening resumes and shortlisting candidates to interview is estimated to take 23 hours of a recruiter’s time for a single hire. Our project aims to solve all the above listed problems. By evaluating all candidates against the same screening standards, the process will be more objective, fair and accurate.

How ML Helps Us?
Machines are better at certain things such as:

1. Sourcing
2. Screening
3. Matching
4. Assessing
   Helps to eliminate human bias, like candidate’s age, race, and gender by assessing candidates purely on their merits. Low-value, time consuming recruiting tasks will become streamlined and automated. Recruiter’s role will become more strategic.

Challenges we ran into

Due to some issues faced while scraping LinkedIn, we decided to scrape data from other major competitors such as Indeed, Github and Frapp which are as well used by most of the people nowadays. Due to a reduced dataset size, the KNN classifier's accuracy was not that great initially. After adding a few more entries to the dataset, we were able to get a better accuracy.