by
Simon Minitzer
(Note: See Simon's BIO on the Contributing Authors’ Page)
"The
User Experience is only as good as the underlying data." Yosi Glick
Yosi Glick has been immersed in building
technology for the entertainment industry for 15 years. He doesn’t believe in ‘Cinderella
stories’, and he doesn’t see his focused vision for Jinni having been sparked
by an ‘aha’ moment. Yosi Glick was persistent.
Glick refused to give in to the notion of why
TV and video guides were always presented in an unsatisfying way to its
customers - he saw the standard grid-style guide as a barrier in the user
experience. Glick made it his mission to break through this barrier. It became
his obsession. His motivation came from his customers, reinforced every time he
asked them if they were happy with the services they were receiving. They would
always respond ‘there is no other way’.
Glick was determined to find a way, and
he did. Glick understood that all the guides were the same because, ‘the user
experience is only as good as the underlying data’. With this in mind, Jinni
was conceived. The accepted way to engage with video content was by broadcast
time and channel or broad ‘genre categorization’ for VOD content. This approach
was not user-focused - and made looking for something good to watch like
browsing the stacks at a library. How do I find what I want? How do I know if
I’ll like it? So instead of zapping through tons of channels or reading through
a long list of titles, it made more sense to Glick to be able to ‘reverse
engineer the intention of the director and the scriptwriter’ so as to better
reflect the ‘mood’ of the video content, therefore making your selections as a
consumer a lot more specific to your mood and interest at that moment. For
example, if a user is in the mood to watch something uplifting and humorous
about a dysfunctional family, the guide should be able to find them exactly the
shows and movies with those characteristics.
Jinni was born. Glick, together with his
co-founders, invented the ‘Entertainment Genome™’
Glick wanted to categorize the data better, and he wanted to understand what
the attributes, or as Glick puts it - ‘the genetic makeup’ of what the video
content was really. Was the movie or TV show ‘sad’ or ‘happy’ or about a
‘dysfunctional family’, or ‘nothing goes right’ or ‘obnoxious boss’. It took
the team a year to identify and understand that there were approximately 2,200
attributes that could describe the video content. On average, 30-60 Genome
attributes are extracted from each show or movie. These attributes form the
basis of the classification, and allow you to select content you have an
interest in watching based on your nuanced mood, and not just a broad genre.
How are the Genes extracted? The patented
Jinni technology cross-references multiple data sources including synopses and
reviews. The Jinni algorithms compute the genes based on understanding of the
meaning of the texts, not just simple keyword identification. The magic sauce
of Jinni is its ability to understand the text and ‘humanize’ it by extracting
the true meaning. For example, Google will index text based on what keywords
appear overtly in the text, but Jinni understands the text the way a human
does. This ‘machine learning’ ability allows Jinni to understand that the movie
‘Little Miss Sunshine’ is about a ‘dysfunctional family’, even though the term
never appears in the synopsis or the review. Jinni understands how to seek out
evidence (certain expressions) that collectively imply that this content has
this particular attribute.
Glick is all about TV and movies. He reflects
on how he assembled his team of experts like the opening scene of the movie 'Oceans
Eleven'. He identified the skills he needed to make his vision come to life,
and literally went on a dating spree. ‘Breakfast, lunch and dinner’ as Glick
puts it, ‘but I always paid for the coffee.’ Glick agonized over how it was
exactly like dating for marriage. ‘Sometimes you could know within a minute of
meeting someone that this wasn’t the guy.’ Glick was persistent, and as he
says,’ I was looking for love.’ Glick attributes this careful planning, a
period of 6 months, as the most critical part of Jinni’s lifecycle. The team
went 18 months without funding, and Glick self-funded sub-contractors to build
the alpha version of Jinni.
Today Jinni is funded by the Startup Factory
and DFJ Tel Aviv, has 30 employees, 2 sales people located in Europe and the US,
with the remainder in Israel. Jinni has 9 tier-1 global Pay TV and online video
operators as paying customers for its unique semantic discovery engine. In
addition, Jinni operates a free website.
Glick has said they have some plans for monetizing the B2C website.
Jinni has, to date, won several prestigious industry awards, including an
Innovation award in 2010 from Cable Labs (Denver) and an Innovation award in
2011 from IBS (Amsterdam).
Jinni recently announced Time Warner Cable
and online video provider, VUDU, a Walmart company, among its latest customers.
In the long term, Glick would like to see all pay TV and video guides powered by Jinni discovery technology.
Glick’s advice to aspiring entrepreneurs is
simple. ‘Be honest with yourself.’ He believes you must accept that you will
receive negative comments, but you should do your best to be objective about
the comments. We should be careful of ‘adapting our reality to fit our
desires.’ So in other words, be wary of rationalizing things to make yourself
feel better. Entrepreneurs should always be challenging the assumptions they are
working with. One always needs to be working in reality.
Jinni wants to get to know you. It wants to
get to know your taste in TV and movies. In a sense, it wants to make your user
experience into a fully personalized ‘siri’-like interaction.
While his team came together like that of 'Oceans
Eleven', their motives are certainly a lot more enviable. They are giving
consumers a way to engage with digital entertainment content in a whole new
way, and that is nothing less than just plain awesome.
il.linkedin.com/in/yglick
www.jinni.com
http://www.facebook.com/Jinnidotcom
Twitter(B2B):@jinnidotcom_PR
http://www.jinni.com/
