Duowingo is a win-win spoken language learning platform that 
provides language speaking practice with the help of native 
speakers in real time. It synchronizes with your learning progress in Duolingo. This concept is inspired by the user flow in Uber and Duolingo.


Nov 2017 | 2 weeks


Individual Project | Saber Zhang


Role | Visual design, Storyboarding
Video production, ideation



Prototyped with Sketch, video made with Premiere and Keynote, modeling by Fusion 360


Objective: Interaction Mashup

Design a new app that combines two non-obviously related existing apps in a way that modifies people’s behavior. The goal is to take the features and micro-interactions from each app that best complement each other to create something uniquely useful: 1 + 1 = 3!

Where are the native speakers?

When I first came to the United States, my English speaking proficiency was below average that I couldn't even have a normal conversation with my American classmates. The lack of verbal practice with native speakers led to my inability to conduct my thoughts fluently in English. As a result, I designed an app that helps people learn second languages with native speakers.

From the given project, I found two potential opportunities from these

two apps:

  • Uber: Most of the Uber drivers I’ve met were immigrants from other countries. This makes for a diverse set of native language speakers.

  • Duolingo: People can't practice their language speaking skills on Duolingo.

 What if Uber combines with Duolingo?



First Attempt:

Let Uber drivers teach passengers their mother tongue.

HOWEVER, Each passenger’s language skills are at different levels. It’s neither efficient nor helpful if the conversation 

between driver and passenger is not coordinated with passenger's language level.

Second Attempt:

With their Uber rides, Duolingo users can request Uber drivers to teach them the language they want to learn.  Uber drivers can see passengers' language levels 
in Duolingo.

HOWEVER, For some languages, it’s difficult to find an abundance of Uber drivers who can speak them.

Sorry, we cannot

find Uber drivers

who speak Nepali

in your area.

After two failed attempts


I revised my plan on how to make native speakers more accessible to language

learners. Instead of just focusing on the Uber drivers' concept, I dove deeper into the core idea of Uber.



 Uber Driver is a native speaker who can provide language speaking services. 


What can I learn from Uber and Duolingo?

I branched from Uber drivers to the focus on how language learners can improve the global community. After deciding on the concept, I analyzed the intention behind both user interface designs and selected what I accumulated from both of Uber and Duolingo to design the new app.

Uber's user interface is very considerable and informative. It ensures the user about everything they need to know before, during and after the riding experience. There is only one action from the user needed per page, which makes the whole ordering experience easy and accessible.

Duolingo turns the entire language learning experience into an online game, while structurally maintaining a set of lessons with increasing difficulties. Duolingo continuously gives users positive feedback on their learning progress, as well as providing a small and exciting test at the end of each session.

Pay a small fee for real-time language practice 

After analyzing Uber and Duolingo, I mainly used Uber's starting page with some of Duolingo's design of giving positive feedback. Here is the final concept model of Duowingo. By paying a small fee for a certain amount of time, users can order the conversation service with a native speaker from the global community in real time. Duowingo synchronizes with the user's learning progress on Duolingo, so the native speaker will know the user's level.

Global Community

Design under the existing structures

The meshing of existing user interfaces and interactions of the two apps sculpted my concept and prioritized the actions I want the user to make on each page. Below are the final user interfaces based on the concept model (animation is shown in the video).

duowingo flow.png

Key interface mashup

Home Page & Order Page

The starting screens are used

for ordering the practice service. I divided the users' actions into two. The two actions that influence the price are time needed for practice and types of practice. Following the flow of Uber's starting pages, I made the duration of practice as the first action and then the type of practice.

Waiting & Conversation Page

To make language speaking practice fun and comfortable, I used the avatar from Duolingo to represent speakers and learners. In the waiting page, I designed a similar experience to those of online games.

Next step

Explore the possibility of execution

I'm curious about whether this idea could meet real needs and if it's executable. I'll make more low-fi prototype to see if people accept this method of language learning transaction.


Our experience is structured in the user interface

By studying different user interfaces and integrating them into my project, I discovered users' mental models of decision making were hugely influenced by how the UI presents the information. From now on, I'll start taking notes from the simple UI I see in the daily life and question the intentions behind it.

The potential opportunities are hidden between seemingly unconnected applications

If we regard the different apps as different systems functioning for different purposes, combining them together smartly can create something totally unexpected. This insight increases my passion for experimenting with different systems.

Going Back To See More 

Archived Works