DEEP

The explosion of content and content creators over the past several years has generated fierce competition to engage and retain users. New platforms, new formats, and the sheer volume of data on a 24-7 basis puts pressure on content creators to tell their stories quicker, with better visuals and interactivity, while at the same time struggling with narrower margins from advertisers and subscriptions. While the pie may have gotten bigger, the size of each slice keeps shrinking, with more and more competition for users’ attention.

Sports fans see a game or a season as a series of dramatic moments. At DEEP, we see a rich, endless tapestry of valuable data to be captured, catalogued, and made accessible whenever a fact reveals a story through context.

Our technology harnesses AI to build and continuously update a vast, responsive database to support the storytelling process. Once we have captured the data, we catalogue it according to our proprietary “lingo” -- the essential building blocks of stories built from the data.

Like LEGOs, each piece of data available through DEEP provides the basis for countless fact-based stories generated by DEEP automatically to fit dynamically changing contexts, or manually through DEEP’s web-based Editor.

The Need

Content creators are challenged with creating an endless flow of content. More than 4 million blogs (http://bit.ly/2OpX1TH ), 2 billion facebook posts (http://bit.ly/2Oqka8u ) and millions of articles are created every single day and the numbers keep growing.

From professional publishers to bloggers and to social sites...more and more stories are generated with more creators involved. At the same time revenue per article goes down as research indicates the average human attention span is now shorter than that of a goldfish.

Eventually for content creation it boils down to few simple facts:

• If it's not visual …. it's transparent

• If it is not interactive....it is boring

• If it can't generate revenues… it's a burden

• If it needs data and research.... it’s not worth the effort

What if we could take the billions of text items created every day and make them visual, engaging and money generating? What if we could do it all automatically? If you are a publisher, blogger or social manager you only need to add 1-line of simple code and voila… your text comes to life with video, matchups, info-cards, polls and more.

Enter DEEP, a service for visual storytelling at scale. A service that can turn any textual article or statistics into an engaging, contextual and visual story, fully automated.

DEEP enables the scalable creation of eye-catching, quality videos, infographics, and interactive widgets especially produced for each story.

Using advanced AI, DEEP continually tracks information so that the service is prepared to rapidly produce content in step with latest developments. This data is packaged into a series of automatically produced infographics, interactive widgets, and videos that are ready to post, providing publishers with a selection of fresh content to plug into their stories – just pick and play.

The most engaging content is a good story. No technology, no matter how sophisticated the AI, can match the human ability to tell a good story. But similarly, no human can match AI for speed in collecting, analyzing, and cataloguing data. WIth DEEP’s ability to analyze and translate text into contextual infographics and videos on the fly, humans focus on storytelling and let technology empower them with the most relevant, contextual facts, updated in real time.

The Technology/How does it work?

DEEP uses smart-machine-learning (AI) to collect data and analyze text to create timely, contextual visual stories. Automatically, in an instant, and at scale.

The slide below shows the main components behind this magic.

DEEP technology has 3 principal components:

1. DEEP knowledge-graph – DEEP’s proprietary AI technology scans the web for data -- anything from simple stats (e.g. Messi scored 26 goals) to interesting do-you-knows (e.g. Ronaldo bought a Ferrari 250 Testa Rossa). As long as it is factual DEEP gets it. Each piece of data is catalogued according to our DEEP Lingo (a kind of digital LEGOs) and stored into our vast database. These digital LEGOs allow for easy and intriguing connections to turn into stories.

2. DEEP Story engine – With the DEEP Lingo, DEEP creates stories in response to a user’s search or automatically, by analyzing text to find context. Like building with LEGOs, DEEP stitches together the most relevant, interesting data and arranges the data into a variety of story formats, including Did-You-Know, matchups, timelines, lists of similar people/events, and game summaries.

3. DEEP Visual composer – DEEP integrates data, images, and format templates to create engaging content ready for publication on the web or social media.

Below is an example of how the same story could be enhanced in 2 different ways, matchup-poll widget and a video:

DEEP’s product is already commercially deployed, providing content creators scalable visual storytelling for sports and celebrities, with plans to expand into finance, politics, and additional sports. In addition, in 2019, DEEP will continue to develop more advanced data insights, more customized automation tied to customers’ KPIs (e.g. engagement, revenue, time-on-page), and additional interactive formats.

Business Model

DEEP employs a two-prong business model to address the different ways in which content creators monetize their content.

• SaaS (subscriptions service) - Content providers subscribe to monthly packages of content created through DEEP (e.g. 1,000 videos/widgets per month). Depending on the extent of data used, subscriptions range from $1,000-$20,000 per month.

• Revenue share - Content providers enter into an agreement with DEEP to split the revenue from ads inserted into content produced with DEEP (e.g. video pre- or mid-roll, widget AdSense in matchup, sponsored videos/widgets). Revenue Sharing agreements grant DEEP between 30%-50% of the total revenue generated by DEEP-produced content.

• While generally one of these two models is dominant, DEEP offers the option of combining the two.

•