Ask blue about water in Az

Project

Az Water Chatbot

Year

2023

The AZ Water Chatbot is a rule-based conversational assistant designed to enhance customer support for water utility services by providing instant responses to common inquiries, reducing call center volume, and improving user satisfaction.

My contributions included conducting user research to identify key pain points, designing structured conversational flows, and creating high-fidelity prototypes in Figma. Through iterative usability testing, I refined interactions to ensure clarity and efficiency. Additionally, I developed a scalable design system for consistency and ensured accessibility compliance across mobile and desktop platforms. This chatbot significantly streamlined customer engagement while reducing operational workload for support teams.


My Role: UX Designer and Researcher

Scope of Work

Web Design
Visual Design
App Design
UX design
User Research
Usability Testing

Overview

The Arizona Water Chatbot was created to give residents quick access to water info, like conservation tips, utility help, and alerts. After user research showed issues with navigation, accessibility, and feature discovery, I used usability testing to identify problem areas.

Feedback indicated users struggled to find critical info, especially in emergencies, and the chatbot didn’t fully meet diverse needs. Based on this, I improved the design by adding personalized, location-based info, chat history for revisiting conversations, and multi-language support for non-English speakers. I also prioritized emergency features for quick access. These updates made the chatbot more intuitive and inclusive across web and mobile.

Research and Discovery

We conducted usability testing and analyzed user feedback on the existing website version, identifying key pain points that hindered the user experience. Through user interviews, task analysis, and surveys, we found that users struggled with complex navigation, poor accessibility due to small font sizes and text-heavy pages, and low feature discoverability, where essential functions like emergency alerts were difficult to find.

This feedback provided valuable insights that directly influenced the redesign, aiming to simplify navigation, enhance accessibility, and ensure critical features were easily accessible to improve overall user engagement and satisfaction.

Problem Statement

❝Users faced challenges navigating the Arizona Water Chatbot, with unclear pathways and limited accessibility features that made it frustrating and less usable, especially for non-English speakers and those with varying abilities.❞

Design Enhancements & Solutions

User Interface & Customization

One of the biggest opportunities we found was improving how the chatbot felt to use, especially during first impressions and repeat visits.

  • Redesigned Splash Screen
    The original landing page didn’t clearly explain what the chatbot could do. I redesigned it to be more welcoming, with clear messaging and visuals that helped set expectations and build trust right from the start.

  • Login & Chat History
    Many users were asking the same questions multiple times. To address this, I added a login feature with chat history, so users could easily revisit past conversations and track previous interactions without starting from scratch.

  • Personalized UI Options
    To make the experience more comfortable, I introduced small but impactful personalization features like font size adjustment, light/dark mode, and layout tweaks, giving users more control over how they interact with the chatbot.

Multi-Platform Accessibility

We wanted users to have a smooth experience no matter where they accessed the chatbot — whether on their phone during a water emergency or on a desktop at work.

  • Designed for Web & Mobile (Mobile Prototype)
    I created responsive designs tailored to both desktop and mobile platforms, ensuring users could easily interact with the chatbot wherever they were — without struggling with layout or usability issues.

  • Consistent Experience Across Devices
    Special attention went into maintaining visual and functional consistency across screen sizes. Buttons, typography, and conversation flows were optimized to feel familiar and intuitive, whether on a big screen or a small one.

Language Support and Inclusivity

Given Arizona’s diverse population, it was important that the chatbot didn’t unintentionally exclude anyone based on language.

  • Added Language Selection
    I introduced a simple language switcher at the start of the conversation, allowing users to choose their preferred language before interacting, making the tool more approachable for non-English speakers.

  • Expanded Access for Multilingual Users
    This change wasn’t just about translation, it was about access. By supporting additional languages, we opened the door for more users to get critical information and support without friction or confusion.

Real-Time Information & Emergency Assistance

In moments of urgency, like a water outage or pipe burst, users need quick, reliable access to help. The original chatbot didn’t support that, so we made it a priority.

  • Location-Based News & Alerts
    I added a dynamic news and alerts section that surfaces local updates based on the user’s location, from scheduled maintenance to boil-water advisories, helping people stay informed without digging through city websites.

  • Emergency Assistance, One Tap Away
    To make the chatbot more action-oriented during emergencies, I designed a dedicated feature that gives users fast access to key contacts: utility helplines, plumbing support, and emergency reporting tools.

Enhanced Resource Accessibility

Users often came looking for links to government services, but finding them in the original chatbot was like digging through a cluttered drawer.

  • Dedicated Resources Tab
    I added a clearly labeled “Resources” section that organizes key links by category, things like billing help, water conservation programs, and city contact pages — so users could find what they needed without guessing where to click.

  • Simplified Navigation
    To reduce frustration and save time, I restructured the information architecture to make resources more visible and logically grouped, improving overall discoverability.

Usability Testing Outcomes

To find the app's usability problems, I conducted Usability Testing with 5 people. The five participants ranged in age from 20 to 35. Key insights were:

Like what you see?
Connect with me

Like what you see?
Connect with me

Like what you see?
Connect with me

Create a free website with Framer, the website builder loved by startups, designers and agencies.