My role on this project: UX Design and Research


In the past two years I have worked to introduce new technologies like AI (artificial intelligence) into physicians tools.


Getting started

I have worked on several AI (artificial intelligence) based solutions in the past two years as my team collaborates with GE, an industry leader in healthcare technology. I have lead design and research for the introduction of features that introduce NLP (Natural Language Processing) to parse unstructured date, advanced imaging processing via CNNs (convolutional Neural Networks) which can interpret medical imaging, and more. You can read more about some of the neat stuff GE is up to here. 

I’ve found it helps immensely for our teams to be informed about the terms and variants of AI; as many people can tell you it is a term used to refer to many things. I made the diagram below as a reference point for categorizing and distinguishing solutions and needing to speak precisely about each.


Our technology, our machines, is part of our humanity. We created them to extend ourselves, and that is what is unique about human beings.
— Ray Kurzweil, American inventor and futurist

I’ve also taken time to learn and share with my coworkers, a few great resources that help to explain the building blocks of AI; like this awesome free online intro, Elements of AI from the University of Helsinki.



I’ve learned much that I can’t share online (with respect to my employer’s privacy and security policies), but I can say that one of the greatest things I’ve started to gather is a sense of how to work in projects like these and how to share knowledge across various roles on the team. It’s not enough for the design team to produce a solution, the knowledge of how and why these tools work the way they do, and the need to bring a unified product to market means the whole team needs a shared understanding. 
For example, one of my co-workers and I created a poster that illustrated some basic approaches to supporting user’s tasks with AI solutions. Materials like this help us have richer conversations: we’re better at saying what we’d like to choose when we know what the choices are!

I have also found that AI features need special considerations when it comes to communications, research and affordance. 
One great tool that I’ve come up with to improve communications is a sort of “cheat sheet” for AI features that are slated for introduction to products. These documents are like a product spec sheet and let the team make sure they’ve communicated the basics of an AI solutions inputs and outputs so that information is not lost as the product passes through various stages of concept, design, development and deployment.


My journey of introducing these technologies to the world continues! I am constantly reading and learning more about what works, what doesn’t and how AI is changing the healthcare domain. 

Further reading

Here are a few great resources I was informed and inspired by recently:


Consciousness in Artificial Intelligence

John Searle, 1hr 10min video

John Searle is the Slusser Professor of Philosophy at the University of California, Berkeley. His Talk at Google is focused on the philosophy of mind and the potential for consciousness in artificial intelligence.

Watch on YouTube


What exactly does a Convolutional Neural Network see?

Convolutional neural network are used to find patterns in an image. You do that by convoluting over an image and looking for vision.

Read on Medium