Undergraduate project: Chatbot for people with epilepsy

Development of the "Eva" chatbot, a technological innovation for disseminating health information, with a focus on conversational fluidity and user experience.

Role

UX Writer | Chatbots

Industry

Computer Science, Health

Duration

1 year (2021)

Step 1. User Research

In the first phase, I carried out extensive research into conversational flows, analyzing users' preferences and difficulties with existing chatbots. From this, I created detailed personas and identified the need for more interactive solutions, applying the "Visibility of system status" heuristic to ensure that users were always informed.

In addition, I studied other chatbots (benchmarking) in the health area, such as "May Health" and "Mila", to map the user experience, identify critical points and apply improvements based on UX practices.

Stage 2. Ideation and Planning

After the usability analysis, I surveyed 74 frequently asked questions about epilepsy in Facebook groups, using the Card Sorting technique to organize the information.

With the personas mapped out, we defined the chatbot's personality, prioritizing empathy and trust, and planned the development using the Lean Canvas, addressing business objectives, user context and conversation strategies.

Step 3. Chatbot Flow Design

I developed the chatbot's flowchart, divided into five areas with pre-defined questions, making it easier to create connections between users and information.

I implemented a suggestion box for cases where the user couldn't find what they were looking for, enriching the database of questions.

Step 1. User Research

Stage 2. Ideation and Planning

After the usability analysis, I surveyed 74 frequently asked questions about epilepsy in Facebook groups, using the Card Sorting technique to organize the information.

Step 3. Chatbot Flow Design

I developed the chatbot's flowchart, divided into five areas with pre-defined questions, making it easier to create connections between users and information.

I implemented a suggestion box for cases where the user couldn't find what they were looking for, enriching the database of questions.

Step 4. Implementation in Landbot

The flow was created on the Landbot platform, where I used variable logic to personalize communication with the user.

When we receive the user's name, we receive it as "@name", saving this variable for a future query or return in a message:

"Hi @name, how are you?" or, "Thanks for testing our chatbot @name...", for example.

Stage 5. Presentation and Final Documentation

I prepared a detailed presentation and documentation, highlighting the importance of interactive design for health education and how the project can positively impact the sector.

The documentation included the entire design journey, from research to the implementation of the chatbot.

Reflections

This project reinforced the importance of continuous user research, iterative design and customization of solutions to meet users' real needs.

These practices have strengthened my collaboration and problem-solving skills, which are essential for my evolution in Product Design.

Talk to "Eva" by clicking here.

Skills used

User Search
Conversational Flows
Personas
Card Sorting
Lean Canvas
Usability heuristics
Interactive Design
UX/UI Design
Product Design
Design Iteration
Chatbot platforms (Landbot)
User Experience Personalization

Reflections

This project reinforced the importance of continuous user research, iterative design and customization of solutions to meet users' real needs.

These practices have strengthened my collaboration and problem-solving skills, which are essential for my evolution in Product Design.

Talk to "Eva" by clicking here.

Skills used

User Search
Conversational Flows
Personas
Card Sorting
Lean Canvas
Usability heuristics
Interactive Design
UX/UI Design
Product Design
Design Iteration
Chatbot platforms (Landbot)
User Experience Personalization

Ana Neves

Ana Neves

Ana Neves