< Disney Connections
Big Data, Dark Data
and Disney
Exploring early Artificial Intelligence with Disney, Deloitte contracted with us to build an intelligent, multi-faceted app that could leverage complex relationships between consumers, content, and commerce. Our goal with the prototype was to prove out the power of graph databases and machine learning in a practical application. Since nothing like this existed at the time, it was easy to take a step back and Think Outside the Screen, but not quite so easy to Simplify.
The Graph - Design Decisions
Visualize the graph.
We chose to visually represent the graph database rather than keep it behind the scenes. The UX led to deep engagement as users explored connections they didn’t know they had.
Utilize motion.
Users could tap, swipe, and drag nodes on the graph to focus on other connections, or simply move the iPad to navigate the graph in 3D space.
Contextual data.
Selecting a node displayed relevant content in the right panel based on user interests, recent activity, or even time of day.
Live Connections - Design Decisions
Use the letterbox.
We utilized the space leftover by letterboxing to show friends, shopping, video details and more. Users could change widgets on the fly and scrub the video to update the context.
Style inspiration.
Using machine learning, we could ingest video, recognize products like clothing or furniture, then find similar items online. Alternatively, we could tag scenes and cross-sell the exact items on screen.
Train the model, one user at a time.
To speed up the machine learning process, as well as strengthen a user’s connections, every recommendation from shows to shopping gave users the opportunity to provide feedback on accuracy and interest.
Outcomes
We were awarded multiple design patents for the user feedback interfaces adding to the startup’s valuation and credibility.
Only a prototype, Disney Connections was never launched, but led to significant investments from both Disney and Deloitte.
< Disney Connections
Big Data, Dark Data
and Disney
Exploring early Artificial Intelligence with Disney, Deloitte contracted with us to build an intelligent, multi-faceted app that could leverage complex relationships between consumers, content, and commerce. Our goal with the prototype was to prove out the power of graph databases and machine learning in a practical application. Since nothing like this existed at the time, it was easy to take a step back and Think Outside the Screen, but not quite so easy to Simplify.
The Graph - Design Decisions
Visualize the graph.
We chose to visually represent the graph database rather than keep it behind the scenes. The UX led to deep engagement as users explored connections they didn’t know they had.
Utilize motion.
Users could tap, swipe, and drag nodes on the graph to focus on other connections, or simply move the iPad to navigate the graph in 3D space.
Contextual data.
Selecting a node displayed relevant content in the right panel based on user interests, recent activity, or even time of day.
Live Connections - Design Decisions
Use the letterbox.
We utilized the space leftover by letterboxing to show friends, shopping, video details and more. Users could change widgets on the fly and scrub the video to update the context.
Style inspiration.
Using machine learning, we could ingest video, recognize products like clothing or furniture, then find similar items online. Alternatively, we could tag scenes and cross-sell the exact items on screen.
Train the model, one user at a time.
To speed up the machine learning process, as well as strengthen a user’s connections, every recommendation from shows to shopping gave users the opportunity to provide feedback on accuracy and interest.
Outcomes
We were awarded multiple design patents for the user feedback interfaces adding to the startup’s valuation and credibility.
Only a prototype, Disney Connections was never launched, but led to significant investments from both Disney and Deloitte.
< Disney Connections
Big Data, Dark Data
and Disney
Exploring early Artificial Intelligence with Disney, Deloitte contracted with us to build an intelligent, multi-faceted app that could leverage complex relationships between consumers, content, and commerce. Our goal with the prototype was to prove out the power of graph databases and machine learning in a practical application. Since nothing like this existed at the time, it was easy to take a step back and Think Outside the Screen, but not quite so easy to Simplify.
The Graph - Design Decisions
Visualize the graph.
We chose to visually represent the graph database rather than keep it behind the scenes. The UX led to deep engagement as users explored connections they didn’t know they had.
Utilize motion.
Users could tap, swipe, and drag nodes on the graph to focus on other connections, or simply move the iPad to navigate the graph in 3D space.
Contextual data.
Selecting a node displayed relevant content in the right panel based on user interests, recent activity, or even time of day.
Live Connections - Design Decisions
Use the letterbox.
We utilized the space leftover by letterboxing to show friends, shopping, show details and more. Users could change widgets on the fly and scrub the video to update the context.
Style inspiration.
Using machine learning, we could ingest video, recognize products like clothing or furniture, then find similar items online. Alternatively, we could tag scenes and cross-sell the exact items on screen.
Train the model, one user at a time.
To speed up the machine learning process, as well as strengthen a user’s connections, every recommendation from shows to shopping gave users the opportunity to provide feedback on accuracy and interest.
Outcomes
We were awarded multiple design patents for the user feedback interfaces adding to the startup’s valuation and credibility.
Only a prototype, Disney Connections was never launched, but led to significant investments from both Disney and Deloitte.