What is CVA?

CVA is a Virtual Agent that builds automated conversation flows for you with never-before-seen scale and efficiency.

Our platform enables crowd-sourced individuals to contribute to building automated conversations by completing simple tasks in a gamified gig-work model platform.

My contribution

Designed the end-to-end experience for CVA's multi-sided enterprise VA web platform.

Brand exploration

Low/high-fidelity wireframes

Prototypes

User testing

Final designs

Design system

Interaction design guidelines

Developed task design systems for crowd-sourced individuals to train client NLUs with great efficiency

Shifted the organization to design-thinking led

Initiated ENG-Design collaboration from conception to completion

Who are we targeting, and how do we reach them?

Sync with Sales Team on target market

Pull competitive research & design reference

Color explorations

Information architecture explorations

High-fidelity explorations

Facilitate design feedback sessions & style selection

Collaborate on strategy for execution

‍

What tasks do our users need to complete?

Identify users and tasks they need to accomplish on a macro and micro level.

Break out screens by key functions relating to key high-level tasks.

Collaborate on UX diagrams with product owner.

Build low-fidelity wireframes.

Collaborate with ENG & Product team on functionality and direction.

Design high-fidelity wireframes.

Ilicit feedback

I built prototypes to help create alignment amongst cross-functional team members, communicate the ideal interaction with our front-end engineer, and test our flows with end-users to continue to refine the experience.

How will our clients manage their VA?

There are a handful of client user types that need to complete tasks associated with their roles. Ex. a Business Leader is probably only looking for high-level metrics to report performance, whereas a Customer Service Lead manages and approves conversation flows.

We put together a user hierarchy to define which screens would be most used by which user, allowing us to adjust complexity where necessary to help each user type to complete their tasks with minimal friction.

Who takes care of managing the top-level client account settings?

The Admin Dashboard was designed primarily for the Client Administrative user, whose main tasks include managing the client's account, internal user permissions, and security.

How will our Crowd Specialists contribute to building automated conversations?

Our Specialist Dashboard enables crowd-sourced individuals to earn cash for completing simple tasks that help train NLUs, among other things related to conversation design. The team developed a performance-based system to incentivize high-quality contributions and eliminate bad actors.

How will our designs remain consistent?

Design systems are important in every industry, but are vital when it comes to environments of high-complexity like the AI space.

I developed our design system to enable our Front-End Engineer to locate all assets, allow for easy consumption to all cross-functional team members, and to serve as a clear guide of our design language to new designers.