AI Voice Filter for Eleven Labs
Abdussalam Popoola
Interaction Designer
Product Designer
UI Designer
Overview
Eleven Labs is an AI audio generator software, that creates realistic voices based on the user's choices.
In this case study, I walk you through my process of improving the voice selection filter to help users;
save time finding the right voice
save credit quotas, thereby leading to better user engagement
The exciting thing is that I’m using Eleven Labs for the first time, so my reactions and expectations were properly documented in video format here;
Problem
As a user, I want to see the character I use in my audio clips to best decide on who i want and not waste my credit quota.
Eleven Labs has a solid AI audio generator. The problem, however, is that it’s quite hard to know what AI character you are using.
Solution
Make the AI voice characters have a face.
Yes, it’s that straightforward.
Here’s why it’s important.
Humans see humans, from their facial expression, race and nationality.
Many assumptions can be pre-determined, making it much easier for users to select the right AI character for their project.
Mock-up example below:
Now let’s talk solutions:
Solution 1 → A face with the country flag
Oh, that’s Alice (she’s cute and from the UK) yippee
Her voice is best used for news and 2 other use cases/roles (good to know)
Solution 2 → Top picks for you
Top picks are a great way to know you got the user's data from the onboarding flow and recommend these voice options to them based on that data.
It enables trust and increases user engagement
source: trust me bro (actually I learnt this from growth.design)
Solution 3 → Search filters
This was what began the whole quest.
You can now filter based on multiple variables so you can get the right voice without using up your credit quota
I know you’d want to see it in action.
I'm open for Product Design Projects like this.
As well as Web Design.
