How a UX designer work in a data science oriented project?

Question

I am a UX designer in a AI research team in financial banking system. We are doing half data science oriented projects and half research work on AI.

Most of our projects are on the credit or client score prediction, financial distress prediction, etc. They are pretty much data science oriented projects. I am the only UX designer in our team. It is interesting to participate in the data science project and learn new things. However, I had zero experience in working with data scientists.

The key difference between a software development project and data science project is that for a software development project, we build things that satisfy the user's needs, and there is always a delivery. However, in data science project, we do not know if it is possible to predict something, because data scientists need to experiment and try to find the insight from the data. Data science solutions do not always come with a delivery.

I found it is relatively easy to identify users from a software development project, because we have specific requirement to make, while it is hard to identify users from a data science project, as sometimes people just want to see if it is possible to do something with data science, the target is so ambiguous.

Most of my work is dashboard design in analytics tools like Tableau, as a UI designer. Sometimes, it is also repetitive, I am kind of fed up with such things. I want to do some more interesting things like user research, to know what users think, how they behave and then experimenting.

Here is my dilemma:

I am interested to work in such a data science team, however, I do not have idea how to work with data scientists and what is my role as a designer in the data science project besides a UI designer.

Does anyone have the experience as mine, and share some of your thoughts or ideas with me, please?

Thanks :)

Answer 1

There are different types of 'data science' projects, so you are definitely not limited to what you are able to do at the moment unless you have been given a very specific set of tasks to do.

You might be surprised to find that data scientist apply a very similar approach to solving problems, in particular with the problem definition and solution validation stages as well as following an iterative process.

The nature of the project will dictate whether you are helping them to create an interface to bridge the gap between hidden data and what they are exposing the end-users to, or perhaps coming up with a dashboard to summary or present the information for a more engaging narrative or decision making process.

Just be aware that data scientist may have some strong views around data visualization, and that it will take some good arguments to help shape the visual design so a critical understanding of the underlying data is essential, but also educating them on aspects of user centred design that they are not as familiar with.