Nova, AI Virtual Assistant

How the chatbot reduced wait times by 80% with 24/7 support, anytime, anywhere.

Role: ​UX / UI Designer

Scope: ​UX / UI Design, Branding, Research, Application Design

Tools: ​Figma, Maze, Miro, Adobe After Effects

Duration: 12 weeks

Jump To:

Goals  ·  Research ​ ·  Persona   ·  Journey Map   ·  Interface   ·  Outcomes     

Context

The future of Customer Service is AI-Human Collaboration!
NOVA - Our virtual digital assistant powered by Generative AI helps customers 24/7 without requiring a human agent!

The Story Behind the Birth of Nova

In the digital era, organisations across all sectors are expected to excel in customer service.  ​Traditionally, customers need to contact the call center during business hours to resolve support issues. Often, there are long wait times or customers have to wait for the center to open, leading to severe dissatisfaction. ​This situation highlighted the need for an AI-powered chatbot that can assist customers 24/7 without requiring a human agent.

The Goals

As the digital technology has become more pervasive, there is a need to reduce the time required to resolve customer issues.

The aim was to design a chatbot which can act as an automated self-service option aimed at enhancing the customer experience and lightening the workload on human staff.  This effectively, helps Businesses improve their Net Promoter Score (NPS).

SIMPLICITY

My main goal was to create a simple and smooth interface for a chatbot that enabled dynamic and meaningful conversations. The aim was to create an interactive and enjoyable experience that would encourage users to return.

ACCESSIBility

Create a chatbot that was accessible to users with different abilities all around the world. For example, include speech-to-text features to help users with visual impairments.

efficient query resolution

I aimed to design a chatbot that could understand user questions, give accurate information, and guide users to solutions efficiently. The goal was to reduce the time it takes to resolve issues, improve user satisfaction, and make the chatbot a trusted source of information.

Positive brand association

I wanted the users to not only see the chatbot as useful, but also as the brand ambassador where the users feel good about the brand whenever they interact with the chatbot.

multiple device support

This virtual agent can be invoked from wide range of devices like desktop, tablet and mobile and would be available 24/7.

Research

I conducted an extensive research to understand the users' feelings, behaviours, aspirations and pain points.  After some stakeholder interviews, I identified the biggest questions and objectives for initial research.

The Biggest Question

Who might use the chatbot and why?

Based on our extensive research, we identified the below cohorts who could benefit from a chatbot.

Small Businesses

Who?

  • Local retail stores
  • Small e-commerce sites
  • Independent real estate agents

Why?

  • Provide efficient customer support with limited resources
  • Compete with larger businesses by offering 24/7 support
  • Reduced operational costs
MID-size businesses

Who?

  • Mid-sized healthcare providers
  • Educational institutions
  • Mid size retailers

Why?

  • Streamline customer interactions
  • Personalise assistance
  • Reduce operational costs while handling growing customer base
large enterprises

Who?

  • Major financial institutions
  • Large e-commerce platforms
  • Telecommunications companies
  • Travel and hospitality firms

Why?

  • Centralised platform to manage customer inquiries across multiple channels including websites, apps
  • Improved response times
  • Enhanced customer experience
  • Gather insights from interactions for better decision making

Definition

Persona

As a part of the project, personas were identified through interviews with representatives from various organization sizes, including small businesses, medium-sized businesses, and large enterprises. By analyzing the unique challenges, goals, and needs expressed during these interviews, three distinct personas were identified. Each persona highlights the specific requirements and expectations for chatbot implementation within their respective contexts.

small businesses
MID-size businesses
large enterprises

We identified large enterprise businesses as the primary persona to target for the initial MVP launch.  

Large enterprises such as financial institutions, large e-commerce platforms, telecommunications companies etc. deal with a diverse range of customers and complex business operations. Chatbots can serve as a centralized platform for managing customer inquiries across multiple channels, including websites, mobile apps, and social media. Our chatbot could deliver value quickly by improving response times, enhancing customer experience, as well as generating key insights for business decisions.

As a result, we decided this as our primary persona as it would help validate and test this solution for a wider spectrum of users.

Customer Journey Map

Once we had identified Large Businesses as our target persona, I conducted a workshop for the team to collaboratively map out the user journey. As the team deliberated each step in detail, some gaps emerged as mentioned below.

User Goals What They Say Expectations Pain Points Opportunities
Call customer support to speak to human agent​ (Traditional Approach) Easily connect to a human agent to resolve their issue. "It takes too long to connect to an agent."

"I keep getting transferred between agents to reach the right support team."
Easily and quickly get the issue resolved with the help from the agent. It takes too long and too many calls to get simple issues resolved at times. Creation of a virtual assistant to resolve issues without having to call customer support
Awareness of chatbot Discover the chatbot and its features "I need a tool that can make my tasks easier." Clear information on how the chatbot can address specific needs. Lack of awareness about the chatbot's existence. Prominent marketing to highlight key features and benefits.
Interest Learn more about the chatbot "How does this chatbot work, and can it save me time?" Access to a brief demo or tutorial showcasing chatbot capabilities. Uncertainty about the chatbot's actual value. Demo video or interactive tutorial during the interest phase.
Consider Assess the chatbot's awareness "Are there success stories or testimonials from users?" Testimonials and case studies that highlight real-world benefits. Doubts about the chatbot's reliability. Showcase success stories and user testimonials prominently.
Decision Decide to try the chatbot "Let's give this chatbot a shot and see how it goes." Clear CTAs for chatbot interaction and introductory offers. Hesitation about committing to a new tool. Hesitation about committing to a new tool.
Onboarding Familiarize themselves with the chatbot "I need to understand how to use this without any hassle." Simple onboarding tutorial with step-by-step instructions. Confusion about chatbot features. Guided onboarding with clear instructions and interactive features.
Engage
Actively use the chatbot for tasks "Let me ask the chatbot for info or assistance." Smooth and interactive chatbot interface for queries. Difficulty in getting desired info. Implement natural language processing for more accurate responses.
Task Completion Complete a task using the chatbot "Great, the chatbot helped me get the info
I needed."
Technical glitches or integration issues. Technical glitches or integration issues. Regularly update and improve integration for a seamless experience.
Feedback Provide feedback on the chatbot experience "I found the chatbot useful, but it can improve in this aspect." In-app survey or feedback form to gather user opinions. Lack of channels to express opinions or provide feedback. Implement user-friendly feedback mechanism.

The Solution

In order to design the best user experience, I have looked at various possible paths the user might take during their journey and examine what happens during each step.

The chatbot is designed to allow users to make queries at any time and receive instant responses.

Interface

Sketching

In the field of design, where time is both precious and a catalyst for creativity, the 8-minute rule emerges as a guiding principle for unleashing the power of hand sketching.  I kicked off sketching the key screens of my chatbot guided by the 8-minute rule. This approach helped me to shape the rapid ideation and conceptualization of my chatbot's conversational interface.

Medium Fidelity Wireframes

As I started this phase of the Design Thinking process, it was essential to create the medium-fidelity wireframes as they serve as a bridge between the low-fidelity and high fidelity wireframes.  They helped to provide enough detail to convey design intent while maintaining a level of abstraction that allows for iterative changes.

Style Guide

Building a style guide was essential to achieve consistency, efficiency, and a unified user experience across various devices. Our aim was to establish a style that not only reflected the brand identity but also streamlined the design and development processes.

I defined a set of core principles that encapsulated the brand values, usability considerations, and a commitment to accessibility.

This involved defining reusable elements such as buttons, forms, typography, and color schemes. Each component was meticulously crafted to ensure consistency and scalability across our digital ecosystem.

The creation of a design system was not a one-time endeavor but an ongoing commitment to improvement. Prototyping played a crucial role in validating our design system. We created interactive prototypes, allowing us to simulate user journeys and identify potential friction points. User testing became an iterative process, enabling us to refine our design components based on real user feedback.

Font - Roboto

Selecting the right font is a key decision in crafting a design system that not only captivates users but also aligns seamlessly with the brand identity. After careful consideration, I chose the Roboto font as the primary font of our design language. ​One of the main reasons for adopting Roboto is its unparalleled versatility.

Designed specifically for digital interfaces, Roboto harmoniously adapts to various screen sizes and resolutions.   Legibility is paramount in delivering a positive user experience, especially in the context of a chatbot where communication is key. Roboto's balanced letterforms and open, modern design contribute to excellent readability, enhancing users' ability to engage with the chatbot effortlessly and understand information at a glance.

Colour - Blue and Grey Shades

Selecting a color for a design system is not just an aesthetic decision; it's a strategic choice that influences user perceptions and interactions.  Blue is universally associated with trust, reliability, and professionalism.

In the context of my chatbot, which serves as a digital companion for users seeking information and assistance, instilling a sense of trust is paramount.

The calming and stable nature of blue contributes to a positive user experience, fostering a reliable connection with the platform.  Blue, with its high contrast against white backgrounds, enhances text readability and ensures that essential information is easily legible, especially for users with accessibility constraints.

Onboarding Screens

These screens welcome the user with a quick introduction, guiding them on how to interact and showcasing key features. They enhance the user's understanding, and foster confidence in the user.

Outcomes

In the dynamic landscape of digital interactions, my chatbot has emerged as guiding light for users through huge amount of conversations and transactions. Beyond its role as a virtual assistant, the chatbot has become a source of insights, providing a detailed understanding of user behaviors, preferences, and the evolving landscape of user needs.

I requested 25 users to test the prototype and share their feedback.  I received interesting responses from users of various different personas.  Large amount of users found that it was very easy to use the chatbot and it took one tenth of the time to get the issue resolved as opposed to spending 20+ minutes to discuss the issue with a human agent at the call centre.

My Other Works

Omnichannel Experience at Three, UK
La Cuisine Royale, Restaurant App
ESSA Health - Responsive Website