Now Reading
How AI Is Helping Viu Close In On Their Game Plan Of Hitting The 100 million User Mark


How AI Is Helping Viu Close In On Their Game Plan Of Hitting The 100 million User Mark


According to a source, the OTT industry in India is estimated to become a $5 billion industry by 2023. Over the last couple of years, Video-On-Demand market in India has gained tremendous popularity with content players such as Netflix, Amazon Prime etc. And this rise in the market has pushed the players to create and produce more original content to build a huge audience base.



Today, there are several OTT players in the market that are competing to become a leading platform and are rapidly tapping on a higher audience base, and Vuclip is one among them. Founded by Nickhil Jakatdar, Vuclip is one of the most emerging platforms in the country right now.

In order to have a clearer view of how Vuclip is successfully marking its presence in the OTT market in India, Analytics India Magazine caught up with Abhijit Bhide, SVP – Engineering, Vuclip.

AIM: Can you tell us about Vuclip’s journey towards the OTT space?

Abhijit: Vuclip has always been in the space of video technology. Before entering the OTT space, Vuclip was one of the leading players in providing video technology and short form content. Vuclip had filed 4 patents and that too before we entered the OTT space. About three and a half to four years back, Vuclip got into the OTT space with long-form content, and now Vuclip or Viu is in 17 countries that are all predominantly emerging nations.

We have a long history in the video optimization as well as providing content in the end-user space overall — working by themselves as well as with partners across the world.

AIM: Keeping the Indian OTT industry in mind, how Viu is playing a vital role in solving the pain points in the industry?

Abhijit: India is an interesting country as well as one of the most competitive from an OTT perspective. How our user consumes in Tamil Nadu is completely different than how the user consumes in Maharashtra. When we had launched in Myanmar and South Africa it was totally two different places, and the journey management there is different. We are noticing the same thing in India.

So, when we are launching in a region, we follow the same philosophy that we used while launching in a new country. We go on a hypothesis based on our ML models, based on what we have learned on similarities of profiling, and then based on the data we collect. We use the collected data for doing our ML model for content creation because Greenlight content is very expensive. So we are working heavily on models that say — what kind of content we should fund, what kind of storyline will work etc. There are 4 to 5 categories of content we have where it is not just the genre, but the hero content — it’s basically what we call as needle mover content. So, based on the data, based on what are the trends and also what we have done with respect to marrying content consumptions with user profiles, we customize and then launched in a particular region.

AIM: As you mentioned journey management, can you tell us more about it?

Abhijit: On the consuming app side we are available to our end users on all form factors — iOS, Android, tablets, TV, Firestick and equivalent. So, the technology in this area is with respect to the consuming app. Then we have ML and recommendation engine as well as engines called user journey management. If you are a nation where bandwidth is very expensive, but still want to consume the best content, everything gets automatically optimized based on the region. Also, there are people from regions who are prone to offers and who don’t care about offers, and in that case, we intelligently and automatically optimise content.

AIM: Talking about the recommendation engine, how much time does Viu’s recommendation engine take to understand the consumer?

Abhijit: We go with the hypothesis, which is based on first-party data and third-party data. First party data takes time to collect compared to third-party data. So based on third-party data as well as what we have learned from the other market, we do profile matching. For instance, if it’s a brand new country we are launching, and we are targeting people between 14 years and 34 years, we try to match the collected data with kind of content and recommendations.

Also, another hypothesis is based on profile data; we do the profiling and ML model shares certain recommendations. Then once we launch it, we start collecting the data and when we reach a certain critical mass, the ML engine becomes more accurate. So basically, what we have seen is, in a brand new place, within a month, we are able to get a fairly good set of recommendations for the end users.

AIM: About data science, would you like to tell us how big is the data science team at Viu And the core areas where the team is currently working on?

See Also

Abhijit: At Viu, we have data science with advanced analytics team and the team size right now is about close to 20.

Talking about the core areas that the team is working, there is data mine, a lot of analytics on terabytes and terabytes of data that is being generated and also, algorithms and heuristics. In terms of content, we are working on Content insights, business insights and consumer insights.

  • Content insight:  We break the content into metadata, and based on the metadata, we keep learning what is getting consumed in what pattern and keep improving our recommendations for the end user.
  • Business insights: It is basically how are Viu is performing. Our user journey management is a great example. It’s an engine and it’s something for which we have filed a patent. It auto segment user experience, collect data, and based on that data takes automatic decisions on what is the next experience to push to them with no manual intervention. So, it drives a lot by the business impact.
  • Consumer insights:  It is all providing content to our users, collecting the data, and understanding the user. We have more than 200 attributes which include dynamic and static attributes and we keep creating more and more attributes.

There are multiple such ML models which we are building and have built. So we keep working towards advancing our ML/AI systems. That is how the team works.

AIM: What makes Viu different or better than its competitors in the market?

Abhijit: The most significant differentiating thing about Viu is the fact that we are not outsourcing anything — all the IP which we have is developed IP in the Pune office. We have intelligently built our platform where we can personalize the experience, learn about the end user and also spend on content very judiciously using a lot of analytics and data. Also, we are operating in 17 emerging nations, which makes our experience very rich. For example, Myanmar just opened up to the world 3 years back and today, we are one of the leaders in Myanmar.

AIM: What’s your and Viu’s plan for the upcoming years?

Abhijit: We are highly ambitious! Our goal here is to produce and deliver great content and go across to regions of the world and increase our user base. We see ourselves growing to more than 100 million users in the next 2 to 3 years and across much more countries. So, I do believe that on the content play as well as the user play, we will try to expand quite rapidly considering the current numbers we have and the growth we have seen in the last couple of years.



Register for our upcoming events:


Enjoyed this story? Join our Telegram group. And be part of an engaging community.


Our annual ranking of Artificial Intelligence Programs in India for 2019 is out. Check here.

Provide your comments below

comments

What's Your Reaction?
Excited
0
Happy
0
In Love
0
Not Sure
0
Silly
0
Scroll To Top