Natural OS


Help smartphone users complete their intended task without distraction.


Artificial Intelligence & Chatbots


14 Weeks


Interaction Designer


The many apps on our homescreen distract us from the initial task at hand.

If we take a look at our phone, our home screen is filled with dozens of apps. Some of them are on multiples pages and some of them are hidden deep within folders. All of the apps on our phone are screaming for attention due to their affordance which can be very distracting and reduce productivity.


Help smartphone users complete their intended task without distraction.

In an AI first world, computers should adapt to how people live their lives. The wall of apps on our phone shouldn’t dictate what we should do. Before unlocking our phone, we have a specific goal in mind that we want to accomplish.


Users are unlocking their phone on average 80 times a day.

Users are constantly checking their phone to find something to do. In a way, the phone is like a refrigerator in that we’re always checking it for something new. We’re constantly digging through apps for new content.



There were 2 main barriers that prevent us from getting our initial task done.


Finding an exact app when you need it the most is a tedious process.

How do you find the exact right app? Is it in a folder or on a different page?


Switching between multiple apps is required to get a task done.

How can apps link with each other in a way that makes sense to the current task at hand?


How do we reduce the number of apps presented to the user?

When I talked to users, finding an app when they needed it most was an issue. Users generally had 2-3 pages of apps; most they rarely use. I decided to ask people what apps they mostly used at certain parts of the day.


We have apps that we always use at specific times and places during the day.

We can reduce the number of apps shown to the user by showing the apps that they frequently use at a specific time and place. We can furthermore simplify the selection by showcasing specific pages or sections from the app because there were cases when only 1 page of an app would be full utilized by users.


How can apps link with each other in a way that makes sense to the current task at hand?

Users mentioned that rather than just showing a selection of apps they often used at the time and place they should be curated in a way that made sense. It was also not different from iOS’s app suggestion feature.

I asked users to cut out sections of apps they used the most and arrange them in order according to a specific task.


You have tasks in mind before you unlock your phone and they involve the use of a selection of apps.

Opening apps randomly from your phone isn’t natural. You have tasks in mind before you unlock your phone and they involve the use of a selection of apps that contribute to that task.


“Stories”, a curated collection of apps that are made for completing a specific task.

The kind of stories shown depend on the time of day and location. These stories reflected the user’s need of using specific apps at specific parts of the day and their habits of using only parts of an app.

Feedback from usertesting

  • The main issue was that I was still using the app drawer to understand the habits of the user. The big question was how do we understand the user’s habits without the app drawer.
  • Users hesitated to interact with the “sections” because they were too used to the traditional full page interface. This new UI paradigm confused users.
  • Users asked how the content in the story would be curated and a lot of technical questions came up like how would you know what section of an app the user used the most.


How do we understand the user’s habits without the app drawer?

I decided to do an experiment on my phone and several other friends’ phones which involved hiding all the apps on my phone. If I needed an app, I would search it using the spotlight feature on iOS. I wanted to understand if the phone was still usable without the wall of apps.

I hid all apps on my phone and searched them whenever I needed them.


Putting search as the forefront allows better understanding of the user’s habits and ensure that every app opened is meaningful.

I noticed several things:

  • You have to decide what app you wanted before you use it.
  • The effort it takes to search for an app confirms that it’s relevant to your current needs.
  • The accuracy of the algorithm will be improved.


Searching was tedious in quick situations like rushing to work in the morning. I realized this was when the stories could come into play. The use of stories became a lot more relevant.


Switching to the Android ecosystem allowed access to a more powerful search bar and better understanding about the user.

For a more accurate understanding of the user’s habits, I believed that switching to the Android ecosystem would benefit the project. Our Gmail account is connected to a wide array of Google products which allows Google to know our daily habits. More importantly, if we’re using search as the forefront of the OS, we need the strongest search engine in the world.

The Google search bar acts as both an app launcher and a search engine.


Returning back to consuming apps in fullscreen.

Users were not used to consuming app content in sections. Futhermore, stacking sections together caused confusion because users thought they were all part of an individual app. I decided to return back to fullscreen apps which users are more familiar with.

How the AI understands the user's habits

The user searches/launches an app through the Google search bar.

The user’s current location and time are recorded.

The time spent on the app is recorded to check its importance.

The next app opened is recorded to make a link between the last one.

How the AI creates a "Story"

The AI finds a group of apps that have a relationship and identifies it as a scenario.

When the user picks up the phone, the AI prepares the right stories based on the current time/place.

The user is presented with stories that are relevant to his or her current situation.

Each story will have a group of apps that will help accomplish the user’s task.


Success is determined by whether users go through an entire story and by the amount of time they spend on their phone (Less time is better).

If users are going through entire stories, that confirms the accuracy of the AI algorithm. If users are spending less time on their phones, the phone is becoming less of a distraction to them.