AI Enabled Employers Recruitment System
AI Enabled Employers Recruitment System

A client came to us with an idea to make the recruitment process much more efficient and simple by implement Machine Learning and Text Analytics technologies into HR market.

The purpose is to help recruiters to be able to focus on interviewing the right candidates rather than to screen thousands of resumes and to get the perfect match. The employer doesn’t need to collect resumes manually from multiple platforms. The only one thing to do is to choose a number of platforms, and the system will collect all resumes and transform them into the readable format.

Entire system is fully flexible and allows to configure publications of vacations and resumes admission based on requirements and financial possibilities.  So that any company has an opportunity to customize its hiring workflow to company's needs.

We were required to build a web app based on this logic. Machine learning allows to screen applicants by their requirements matching level using the matching engine. While text analytics technology automatically converts job descriptions text into data and applicant resumes files (Word, PDF and etc.) into searchable data. App also uses API connections to publish the job openings on multiple job boards. The implemented algorithm learns the employer’s patterns of recruitment and improving the matching results based on that.

Technical solution

Our solution was a multi-user high load application with the real-time data updates, We’ve created a Kanban-like system that allowed to manage resumes. Application required set of user roles with different rights to access data and perform actions. In order to prohibit access to data and / or transactions to those users whose access level is insufficient we introduced security measures. Based on required app logic we did the integration with several systems of resumes recognition and vacancies description. The most appropriate algorithm for processing the file is automatically selected based on the language (Hebrew or English) and file type (resume or job description). In order to make the screening process easier we implemented the algorithm that automatically determine the candidate's compliance with the vacancy on a scale from 0 to 10 points. This algorithm allows to reject all candidates with the low level of compliance. The issue how to store all this amount of data was solved due to integration with two databases MongoDB and Neo4j.

 

Technologies

  • React
  • Typescript
  • NodeJS
  • LoopbackJS
  • MongoDB