Kalibrr’s Artificially Intelligent Job Recommendations Engine is now live
We’re excited to announce that our Data Science team launched the new Jobs Recommendation Engine, Kalibrr’s Artificially Intelligent job matching algorithm, last Monday, March 28 2016.
Our lead Data Scientist, Aivin, shares to us how the new recommendations engine works and why it matters to jobseekers and recruiters alike.
Side note: For budding data scientist and other enthusiast, watch out for a technical blog post by Aivin and his team on our product blog!
TOPICS
How does Kalibrr’s A.I powered recommendations engine work on Kalibrr?
Recommended Jobs Tab
We recommend jobs to you based on your interactions on the platform and the information you provide on your Kalibrr profile.
Related Jobs
At the end of every job post, we recommend jobseekers to look at other related jobs based on the job post’s content, information and how other jobseekers interacted with these jobs.
There are 2 types of algorithms at work here.
1. Relationship Matching:
The A.I learns the relationship between specific jobs based on jobseeker interactions with job posts and correspondingly their profiles.
- It looks at all the available jobseeker profiles on the platform and builds a network of related jobs based on jobseeker interactions with these jobs.
Before, when a jobseeker declares he’s interested in web developer jobs, he will only be able to receive web developer recommendations.
Now, when a jobseeker declares interest in web developer positions, he will be able to receive job recommendations that belong to the same cluster: Software Engineering, Java Developer, Ruby on Rails Developer, etc.
Why it matters for recruiters:
- This is likely to improve the impressions and exposure of your jobs on the platform as jobseekers are able to discover your jobs better.
Why it matters for jobseekers:
- Kalibrr will be able to help you discover new jobs that are a great match for you better than before.
2. Content based matching
Job Recommendations based on content and properties of Job Post.
Why it matters for recruiters:
- Recruiters who create good quality job posts and put more information are more likely to be grouped with other good quality job posts.
Why it matters for jobseekers:
- This algorithm solves the cold start issue: How do recommend jobs to new jobseekers and candidates who haven’t filled in their profile or taken assessments.
- We can still intelligently recommend jobs to you based on the current jobs you are viewing. That means you’ll be able to discover more relevant jobs to you. Awesome!
Bonus Content:
As part of our Jobs Recommendations launch, Aivin and his team built a twitter bot: Kalibot!
Stay tuned for more product updates!
Much Love,
Kalibrr Team.
No comment available yet!