What if you could create a behavioral profile of someone just using their voice in under 45 seconds? That's what we wanted to build with VoiceSignals.
When we started building VoiceSignals we wanted to leverage the power of AI/ML to build a platform that could create personality assessments using just your voice. Your voice has a lot of information in it, more than people realize.
We had an engine that could identify and listen to a Trait Library of 30 Personality Traits/ 8 Mindsets/ 8 Emotions but how would those data points be helpful?
We had three goals:
Build models that can listen for 4 unique buyer types and 6 personality types
Build content that would allow a user to understand and communicate more effectively based on our groupings
Enable the platform to listen for personas that identify target and non-target behaviors. (ei. high-propensity for converting on a lead, likely to default)
Wireframing
We needed to move quickly on these because the clock was ticking and we wanted to get MVP live for beta users to provide input. We chose to do very low-fidelity wireframes to give direction on each module that would be on our screens.
Working with the developers we built baseline conceptual guardrails that allowed them to have containers ready for when the high-fidelity was ready.
I wanted the interface to feel both accessible and appealing, that the attributes and traits would each have their own uniqueness but feel part of the entire ecosystem.
Leaning into the atomic design system developed by Brad Frost we looked at our core atoms and molecules and then built out the larger molecules and various components that we would need. We wanted to target our desktop and tablet viewports which made it easier to not spend too much time on the mobile since this wouldn't be for anything smaller then 840 width.
Final Results
We had built a great working environment and a pretty powerful platform. We had given 50+ demos with several paying customers. But in the end we didn't have enough money to keep the company afloat.
We should have stayed on a simple couple of interactions and stayed focused on the use cases. We tried to 'boil the ocean' and could have been more focused on simplicity.
While we had great initial feedback and people loved where we were going in the end we just went to big. I'm still super proud of all the work. Definitely a learning experience.
