Watson Hello
Giving communication a new voice through a ground-breaking piece of technology.
Objective:
Build an application that utilizes the capabilities of IBM Watson's Language Processing.
The objective the team was given from the main stakeholders was this: A consumer can communicate in-person with non-native speakers with their mobile device at the speed of thought, and give Watson feedback along the way.
My role:
Product Designer / UX Designer — Working along side a team of User Researchers, UX Designers, and Developers.
Research
We began our project by doing some research of our own into the problems at hand, which led us to an array of assumptions we could test and validate.
From here, we were able to efficiently and effectively organize our research plan and questions to ask our Subject Matter Experts as we look to discover, validate, and understand the user’s pain points.
User Interviews/Personas
To seek additional information we went straight to potential users of the application, all with differing background and experiences with technology. This allowed us to cull a range of real user opinions and give them the opportunity to help validate or negate some of our assumptions.
Within a week we interviewed 7 people—travellers, Peace Corps volunteers, a Doctors Without Borders worker, a nurse that frequently had patients that did not speak english, and employees at multinational corporations who had to communicate in a non-native language daily.
From there, we mapped our user personas and user scenarios, providing us with a roadmap of sorts that allowed us to easily identify user pain points and user success along the journey.
Insights
From our learnings, we developed insights to help guide our design work.
Concepting
With our research to guide us, we were bale to generate needs statements which represented a user with a problem with a clear outcome.
e.g. “Marion, the knowledge worker, needs to feel confident in the validity of a translation so that she can keep her focus on the task at hand”
While concepting, we distilled out ideas into easily digestible metaphors of what we were to build. Such as “A Tomagachi for language translation.” or “A Spotify playlist of vocabulary words.”
Rapid Prototyping
While rationalizing our design choices, we were able to rapidly prototype and test user flows in a fast and cheap way by using paper prototypes, this allowed us to test design assumptions quickly as to learn quickly before investing too much into design on a larger scale.
Design
With our ideas and concepts rationalized, it was time to build the product. Working with our SME’s and stakeholders we could test design decisions and feasibility in real time through constant collaboration.
Users can simply activate Watson and it will translate a two-sided conversation in real time.
While communicating through the app, Watson will highlight words without direct translations and link to an outside database, like Wikimedia, to provide more context. These smart words can then be stored in a library to refer to later.
An example of a smart word deep-dive and the information it can provide for the user.
If speech is not an option, users can also utilize a text-to-text translation service.
Reflection
Upon completion of the project, we agreed our MVP would be successful and we could go to launch. From here we decided the next round of refinements would include more in depth privacy settings, exploring the use of technology without internet or cellular connection (localized databases), and and continuing to train the tool to learn and become even more efficient and reliable than it currently is.