Virtual Assistant for Santander Consumer Finance

Development of a chatbot for Santander consumer finance

Role

Product designer

Deliverables

Designs + Prototypes

Flows

Participation in user testing

Technologies

Android

iOS

Angular

Problem statement

Chatbots are each day more present in our lives. We tend to relegate them to solve easy tasks, but technology is advancing at light speed and artificial intelligence gets day by day more sophisticated.

Chatbot applications streamline interactions between people and services, enhancing customer experience, offering companies new opportunities to improve customer engagement.

The idea that Santander Consumer Finance had was introducing a way to help their users actively perform actions they could do manually using the app by just asking the chatbot, e.g. making payments, consulting balance, canceling and even hiring products.

Screens showing the initial flow of virtual assistant

Execution

The market:

The Santander virtual assistant was intended to launch at a global level as part of one of the modules developed by Globile project. As part of the different products that were developed simultaneously, the chatbot was the first one to be released and used.

It responded to the urgent and specific need to help users carry out simple tasks that were made to be really difficult due to the very complex and old UI structures used in the digital product across Santander’s Network.

Research was crucial to understand how similar products where used and their capabilities and limitations.

a phone showing a conversation of virtual assistant

Functionality:

The chatbot was developed in all 3 technologies: iOS, Android and web. We had to understand the limitations of each technology and the expectations for the user of each platform.

We gathered information to understand the most performed actions by our users with the banking platforms and where there existed the most friction.

We tracked the user testings performed by the research team to ensure that our solutions were understood by end users making sure we were improving the experience and delivering successful solutions.

Composition of different virtual assistant screens

Dark mode:

The Global Design System was being created at the time and a novel product we got to use in a beta state in the Globile project. I was the first one to introduce the dark mode as part of an experiment to make sure the automatic translation of the tokens would behave well in my product.

I wrote a report that I shared with the global team about all the risks I encountered with examples and benchmarking and suggestions to improve.

The document was studied by the management team and was introduced to carry out changes that improved the global Design System and the dark mode definition.

Image showing examples of virtual assistant in dark mode
comparison between virtual assistant screens in light and dark mode

This project inspired some articles you can find below:

Dark mode