A.I. — The Winning Factor for the Personalization Battle!

A.I. — The Winning Factor for the Personalization Battle!

The outline of the article:

  • About
  • Which Show/Movie Should I Watch?
  • A.I. — The Winning Factor for the Personalization Battle!
  • How Traditional Recommendation Platform Works
  • How Netflix Achieves this
  • Some More Important Applications of AI
  • Personalized Job Search Experience
  • About glasssquid.io
  • Conclusion


According to reports, more than 80 percent of the TV shows and movies people watch on Netflix are discovered through the platform’s recommendation system.

In this article, you’ll see how Netflix achieves this. Further, you’ll get to know that how the same A.I. based techniques can be used to develop an intelligent A.I. based job matching/recommendation system.

glasssquid.io is such a job search portal that provides personalized job recommendations to job seekers.

In this article, we will see how glasssquid.io is utilizing cutting edge A.I. techniques like NLP (Natural Language Processing) to simplify the task of recommending the most suitable jobs to you.

Machine learning, algorithms and creativity.

Image for post

Photo by Thibault Penin on Unsplash

Which Show/Movie Should I Watch?

This is the question that pops into your mind once you are back home from the office and sitting in front of the TV.

Today, everyone wants an intelligent streaming platform that can understand their preferences and tastes.

From Netflix to Amazon Prime — Personalized Recommendation Systems are gaining importance as they directly interact (usually behind the scenes) with users every day.

A.I. — The Winning Factor for the Personalization Battle!

If you are a Netflix user you might have noticed that the platform shows really precise genres like Romantic Dramas where the leading character is left-handed.

Image for post

Photo by Andy Kelly on Unsplash

How does Netflix come up with such precise genres for its 100 million-plus subscriber base?

It’s machine learning, AI, and the creativity behind the scenes.

Machine learning and data science help Netflix personalize the experience for you based on your history of picking shows to watch.

With thousands of TV shows and movies on Netflix, how do you chose what to watch?

The simple answer is that Netflix offers you a hand-picked selection right to your screen.

How Traditional Recommendation Platform Works:

In the traditional recommendation platforms, they use simple keyword matching.

For example, if you are searching for the “Action” genre, the traditional recommendation platform will use simple keyword matching.

Image for post

Photo by Krists Luhaers on Unsplash

It’ll give you all action movies/shows, completely ignoring your personal interest/watching history related to it.

There are many sub-genre in the action category: Straight Action, Adventure, Action-Adventure, Heist Films, Detective Mysteries, War Movies, Sci-Fi Action, Martial Arts, etc.

By now, you must’ve got my point. I want to tell you that traditional recommendation engines will just pull out everything you ask instead of what you want.

So, It’ll not be easy for you to filter out your choice.

Image for post
Photo by Franki Chamaki on Unsplash

How Netflix Achieves this:

It does this by using a variety of factors. These include:

  • How you interact with their service (like your viewing history, search queries, and personal ratings of content).
  • Data collected from other members on the site with similar interests to your own.
  • Linking all that to information about the titles, such as their genre, categories, actors, release year, etc. over their content.

So, whatever they will recommend you, it’ll be better to date (As it considers your watch history till the moment.)

Some More Important Applications of A.I.:

So far so good. Surely! Entertainment is an essential part of life. But what about the other aspects?

  • Job Search.
  • E-commerce (e.g. Amazon.com)
  • Content Recommendations (e.g. Medium Blog-posts)

Personalized Job Search Experience:

There are hundreds of job search sites on the web, but they’re not all created equally.

Job hunters today need a site that will maximize their chances of finding a position, and that won’t waste their time with old job listings or functionalities that are less than user friendly.

Image for post

Photo by Marius Ciocirlan on Unsplash

About glasssquid.io

This tech-only job search site brings personalized job postings and skills center — complete user-friendly experience — to those in the tech and cyber-security field looking to find new opportunities.

Shows the right jobs to every person instantly.

Candidates can upload a resume and instantly match relevant jobs according to their skills, experience, & interests — and apply for the job.

The system takes into account:

  • The skill-set
  • Job Function
  • Employment type
  • Experience level
  • Domain Expertise
  • Educational qualifications
  • Location and Work Authorization

We have trained our model on hundreds of millions of jobs to encode the natural language of the documents in a highly nuanced mathematical vector space, Here is a visualization of the keywords used in the job description/resume.

Each dot represents one keyword, each keyword is a data point (with 100 dimensions in this case).

Image for post

Using a technique t-SNE, the computer clusters keywords together which it considers related to each other.

t-SNE is a machine learning algorithm for data visualization, which is based on a nonlinear dimensionality reduction technique.

t-SNE is looking for a new data representation where the neighborhood relations are preserved.

The clusters are based on the meaning of the keywords, which make use of word embeddings.

Take again an example of movie genres. If we run the algorithm on those, it’ll form a cluster for various types of action movies, another cluster for comedy movies, a different cluster for romance movies, and so on.

word embeddings or word Vectors are numerical representations of semantic similarities between words

The below image is the t-SNE representation of word vectors in 3-dimensional space and you can see that semantic similarity between HTML5, CSS 3, and jQuery has been captured.

If you want to visualize the same, stay tuned for our next blog-post.

Image for post

Then glasssquid.io, which is AI-Powered Job Search Engine/Portal, will recommend you jobs related to software engineer profession (Of-course with the specific specialization which you have, like Python/Java/Cyber-security, etc.)

Your skill-set e.g. Python Developer with Snowflake tool experience, then the jobs with:

  • Python Developer which has Snowflake as a requirement will have higher relevance in the job recommendations rather than just Python Developer jobs without Snowflake in the job description.

glasssquid.io goes beyond simple keyword matching and uses semantic matching.

Semantic similarity between Job function and your skills provides the best matching jobs.

e.g. If a Python/.Net/Java Developer or a Project Manager resume has some sort of experience in Trading applications or onboarding applications, the jobs which have the expertise required in those domain experience are presented in the higher relevance than others.

The Transition from one job to next:

Suppose 2 years ago, you were a teacher. Then you changed your career to software engineer.

Traditional job search engines/boards will still display you the jobs related to teaching.

However, glasssquid.io takes into account your most recent job role and gives you relevant recommendations related to the software engineering field.

Image for post


Data science in our daily lives gives us access to educated choices and more curated experiences. The computer power and know-how of data scientist results in seamlessly accurate decision-making. Think about that during your next job search!

What’s next for you?

Leave a Reply