FI_banner_button.png

Financial Assistant

 
 

How do we communicate to a customer through chat and voice bots?

Chat bots are all the rage, but there is a lot that goes into them, especially once you add voice. Many of the brand voice guidelines and themes are marketing-based and not suited for one on one conversation with a customer. 

 
 
 

1. My role

I created the user experience flows, UI and visual design elements, screen layouts, text phrasing (uderances) and conversations design.

Goal

  • Make an engaging experience with a chat bot

  • Integrate voice technology

  • Design a financial assistant with machine learning

 
 
 
 

2. The planning phase

Leadership wanted to show how a large financial institution could create a data driven machine learning AI to assist users in their finances and goals over time. 

 I worked with the development team to explore the capabilities and limitations so we could build out a proof of technology.

 
 

Research

I started looking at all the “assistant” apps and technology, especially ones that use voice. I did deep investigation into Siri, Google assistant (including google home), Alexa, and Cortana. I documented how these 4 leaders interact with the user, especially how the bot talks to the customer and addresses itself. I took note of all visual queues these assistants use and started to develop how the project’s bot should “speak.”

 
 

3. The design phase

Working with the business analyst and development team, we brainstormed use cases and scenarios to explore. We focused on 2 themes we’d seen as high desirability in user research: payment and goals.

From there I developed draft conversations for the scenarios. Starting with happy path, I listed:

  • Utterances (words the AI would recognize the user saying)

  • Intents (where the words are supposed to take the user)

  • Entities (variables that allow the AI to surface the correct data)

  • Chat bubble leads and other UI elements

Then we mapped out the branching paths and options to test. This all built into a larger story for the team to present to Leadership and other line of businesses.

 
The AI can help the user find savings.
 

This chat bot surfaces relevant information, like goals, recent activity and progress.

 
 

4. Implementation Phase

The biggest challenge was finding a good voice and technology for the bot’s personality. The dev team started with the native Siri kit, but it wasn’t giving the level of human-esk pronunciation and pause.  I researched the many options before landing on 3 acceptable solutions. I presented these to development for integration testing. 

 
 
 

Closing

This project will provide a great deal of foundation to bots and customer interactions in this technological space. The app and research has been socialized to many line of businesses and sparked more conversations. There are many aspects that have not been flushed out and are now on designer's radar.