Artificial Intelligence Solution for Job Recommendations

Anjali Patiyal
Using artificial intelligence and machine learning for job recommendations to job seekers

Client
An early stage start-up in Europe

Industry
Digital

The Situation

One of the early stage and fast-growing start-ups in Europe was looking for an Artificial Intelligence (AI) solution to solve some of the key challenges faced by Job Seekers and Employers today:

  • Limited analytics implemented by current job portals to filter out relevant resumes for Employers and suitable job openings for Job Seekers.
  • Not high-lighting achievements of job seekers along with related recommendations. Generic recommendations are not very helpful.
  • Replacing a full-time worker can cost up to twice the employee’s salary.

JKT Solution

Based on our experience in Digital transformation, JK Technosoft proposed an artificial intelligence (AI) solution with web and mobile applications development to build the complete platform for both Job Seekers and Employers. The platform also had a robust analytics and reporting module.

Our Digital team evaluated various artificial intelligence/machine learning frameworks like TensorFlow, Caffe, Amazon Machine Learning and Spark MLib. While the implementation is fully service-oriented and can be used with any Machine Learning frameworks after minor changes, JKT has chosen Spark MLib framework which is deployed on Apache PredictionIO platform for building this solution.

Key Features

  • Parsing resume from different languages using Natural Language Processing (NLP).
  • Content based Job recommendation using machine learning.
  • Jobs are recommended based on assessment score.
  • Sentimental and behavior analytics from user assessment.
  • Robust and scalable back-end.
  • Omni-channel availability – Web & Mobile.

Business Benefits

  • Job seekers will get more relevant and focused job opening as per their requirements.
  • Employers will not be flooded with resumes. They will receive targeted job applications.
  • Platform will provide insightful response to job seekers, employers and to the industry
  • It is a cloud-based solution and available on both web and mobile applications for ease of access and use.

Key Components

  • Robust Recommendation Engine Platform, that can be used to deploy multiple predictive engines as webservices.
  • It’s a full machine learning stack, bundled with Apache Spark, HBase and Elasticsearch.
  • support machine learning and data processing libraries such as Spark MLLib and OpenNLP.
  • Batch or in real-time predictive analytics.

Tools/Technologies

  • OAuth 2.0
  • Spring REST services.
  • Apache PredictionIO Platform.
  • Spark, Spark MLib for Machine learning and computation.
  • Data Storage using Postgress and MongoDB.
  • GATE NLP for resume parsing.