Home

Work

Discord, B2B2C

2025

Crocs, B2C

2025

Bank of America, B2B

2025

▸ DigitasAI, B2C

2024

Contact

LinkedIn

Email

I'm glad you're here
Me alegro que estés aca

©️ All rights reserved, 2025

DigitasAI, Empowering users to create agents

DigitasAI, Empowering users to create agents

B2C, Product Design, 2024

I led the redesign of DigitasAI's Add Agent feature to empower users make customized AI agents more successfully. I conducted mixed methods research, created prototypes, ran tests, gained alignment with stakeholders, designed an improved self-serve AI creation tool, as well as partnered with engineers and QA to implement the final product.

The team was a lead designer (me), a product manager, a lead engineer, and a QA specialist.

Go to:

Go to:

Problem

Solution

Outcome

Research

Define

Journey

Test

Overview

Context & problem

DigitasAI is an internal platform with powerful AI tools to help improve people's workflows and outputs. One of the main ways people use the platform is by creating and interacting with their own specialized AI agents via the Add Agent feature. When I joined the team, non-technical users found this feature to be intimidating and confusing, resulting in low usage and a high number of support tickets.

Goal

My goal was to improve the Add Agent feature for increased adoption and task success by:


  • Creating visible steps to help users see their progress in the creation process

  • Adding more context and visual queues to coach users through the process

Solution: Redesigned Add Agent feature

Through multiple rounds of iteration, critiques, and tests, the redesigned Add Agent feature is an intuitive multi-step experience that coaches users through the agent creation process.

Redesigned agent creation flow

Tooltips and instructional text increased user comprehension around the agent creation process.

Tooltips and instructional text

Outcome

  • Increased adoption and task success

  • Reduced support tickets related to the Add Agent feature

Process

Research and discovery

I conducted mixed-methods research to understand why non-technical users were experiencing friction with the existing experience. I conducted stakeholder and user interviews, as well as secondary research to better understand the space.

Key findings include:

  • Long single-screen interface felt daunting

  • Users did not clearly understand the steps necessary to create an agent

  • Non-technical users needed education around technical terms

  • Agent creation was unclear due to no completion feedback

I conducted mixed-methods research to understand why non-technical users were experiencing friction with the existing experience. I conducted stakeholder and user interviews, as well as secondary research to better understand the space.

Key findings include:

  • Long single-screen interface felt daunting

  • Users did not clearly understand the steps necessary to create an agent

  • Non-technical users needed education around technical terms

  • Agent creation was unclear due to no completion feedback

Existing experience and interview snippets

Problem definition

I created a problem statement and goal to drive consensus with my stakeholders on the focus of the project.


Problem statement: Non-technical users struggle to create AI agents due to an intimidating single-screen interface that lacks clear steps, educational context, and user friendly language, resulting in low adoption and task abandonment.


Goal: Create an intuitive, guided agent creation experience that empowers non-technical users to successfully build custom AI agents through clear education, progressive disclosure, and user-friendly language.

I created a problem statement and goal to drive consensus with my stakeholders on the focus of the project.


Problem statement: Non-technical users struggle to create AI agents due to an intimidating single-screen interface that lacks clear steps, educational context, and user friendly language, resulting in low adoption and task abandonment.


Goal: Create an intuitive, guided agent creation experience that empowers non-technical users to successfully build custom AI agents through clear education, progressive disclosure, and user-friendly language.

I created a problem statement and goal to drive consensus with my stakeholders on the focus of the project.


Problem statement: Non-technical users struggle to create AI agents due to an intimidating single-screen interface that lacks clear steps, educational context, and user friendly language, resulting in low adoption and task abandonment.


Goal: Create an intuitive, guided agent creation experience that empowers non-technical users to successfully build custom AI agents through clear education, progressive disclosure, and user-friendly language.

I created a problem statement and goal to drive consensus with my stakeholders on the focus of the project.


Problem statement: Non-technical users struggle to create AI agents due to an intimidating single-screen interface that lacks clear steps, educational context, and user friendly language, resulting in low adoption and task abandonment.


Goal: Create an intuitive, guided agent creation experience that empowers non-technical users to successfully build custom AI agents through clear education, progressive disclosure, and user-friendly language.

I created a problem statement and goal to drive consensus with my stakeholders on the focus of the project.


Problem statement: Non-technical users struggle to create AI agents due to an intimidating single-screen interface that lacks clear steps, educational context, and user friendly language, resulting in low adoption and task abandonment.


Goal: Create an intuitive, guided agent creation experience that empowers non-technical users to successfully build custom AI agents through clear education, progressive disclosure, and user-friendly language.

User journey

I mapped a user’s journey to creating an AI agent, which helped prioritize the information we could show at each step. This also highlighted the opportunity to streamline the journey when a user does not add an index.

I mapped a user’s journey to creating an AI agent, which helped prioritize the information we could show at each step. This also highlighted the opportunity to streamline the journey when a user does not add an index.

User journeys mapped with and without adding an index

Wireframes and user testing

I designed the pages using existing components to highlight system status and included tool tips for a in-depth explainer. I conducted user tests with 10 people to understand how they interacted with these designs and to test comprehension of the agent creation process. A few insights:

  • Multi-step flow and progressive disclosure increased comprehension

  • Tooltips provided helpful context, appreciated by non-technical users

  • Visible system status gave users clarity into the creation process

Overall, the designs increased comprehension around the agent creation process, and we addressed the points of confusion through clarifying the steps to creating an agent with progress indicators, tooltips, and success feedback.

I designed the pages using existing components to highlight system status and included tool tips for a in-depth explainer. I conducted user tests with 10 people to understand how they interacted with these designs and to test comprehension of the agent creation process. A few insights:

  • Multi-step flow and progressive disclosure increased comprehension

  • Tooltips provided helpful context, appreciated by non-technical users

  • Visible system status gave users clarity into the creation process

Overall, the designs increased comprehension around the agent creation process, and we addressed the points of confusion through clarifying the steps to creating an agent with progress indicators, tooltips, and success feedback.

Agent creation designs

Back to home

Next: Discord, Lobbies Integration

Home

Home