Enhancing Healthcare Access: Sydney Chatbot
Introduction
This chatbot is designed to engage with users in natural language conversations to gather relevant health information, offer medical advice, and guide users to appropriate healthcare resources. The chatbot utilizes artificial intelligence and machine learning technologies to understand user queries and provide relevant responses.
Problems and Challenges
Currently all healthcare related data, service actions and member and provider details are present in various corners of an existing company application. The company is in need of a conversational guiding tool which can gather relevant health information, offer medical advice, and guide users to appropriate healthcare resources.
The chatbot should be able to understand and respond to user queries, perform tasks, provide information, and simulate human-like conversations in a seamless and intuitive manner
Client
Elevance Health
Industry
Healthcare
Duration
4 weeks
Roles
Sole UX Designer
The Solution
The design team adhered to a User-Centered Design approach throughout the process, resulting in a chatbot that is intuitive, clean, simple, and delivers quick, effective search results by continuously learning and adapting to users' needs.
​
Most importantly, the final product enhances Elevance Health's ability to save time for both users and customer service agents while boosting user satisfaction.
Timeline
The timeline encompassed research, ideation, design, development, and continuous improvement, ensuring a user-centered approach to deliver a seamless, efficient, and adaptive conversational experience.
1
2
3
4
Discovery
Define
Ideate
Design
5
Takeaway
Discovery
Through my secondary research, I realised that:
-
User expect AI chatbots to understand natural language, perform tasks, and engage in human-like conversations.
​
-
User expect quick responses on information and support for medical inquiries.
​
-
User trust is paramount, as users are concerned about the accuracy and privacy of healthcare-related information.
​
-
Users value chatbots that can assess symptoms and provide preliminary advice and clear disclaimers.
​
-
Adherence to healthcare regulations, such as HIPAA is crucial.
​
-
The chatbot should play a role in educating users about health topics, preventive care, and healthy lifestyle choices.
​
-
Understanding the demographics of the target audience is essential for tailoring the chatbot's responses and features effectively.
​
-
The chatbot should communicate in a friendly and empathetic tone while using plain language to ensure accessibility to a wide range of users.
​
-
Consider partnerships with healthcare providers or institutions to enhance the chatbot's credibility and access to medical expertise.
Popular Medical AI chatbot in the market
Digital Survey
After the secondary research, I conducted a digital survey with our user group to learn more about how they feel about using the existing application and resources. I organised the necessary resources, got in contact with the user group and got 100 pieces of valuable feedback.
Define
With all the notes, documents and data from the digital survey, I put everything in perspective, synthesized with the participants, and created the major problems areas, personas, user experience maps, and scenarios.
Problem Areas
Thorough analysis of the digital survey led to identifying the major problem areas.
I couldn't find an answer to my question. Build your website from a consumers view, someone who doesn't deal with healthcare every day.
- Anonymous User 1
Friendly little chat is not very helpful nor fast. no help at all with anything that's out of the norm... which it seems most of my things are. Had to resort to actually calling and got stuck in that long winded verbal barrage before I could talk to a real person. ARRGGH!
- Anonymous User 2
- Anonymous User 3
User Persona
After a deeper look into the digital survey and problem areas, I created four personas,
Emily
Wants to access information related to her health problems and is open to scheduling a virtual consultation with a healthcare professional.
Maria
Wants to locate a provider specialist who can provide a thorough evaluation of her digestive issues. She also wants to use the chatbot to help her schedule an appointment with the chosen specialist at a time that fits her busy teaching schedule.
John
Wants to gain a clear understanding of the terms, coverage, and benefits offered by his health insurance policy. And he also is seeking guidance on how to navigate the process of filing claims and seeking reimbursement for medical expenses covered by his insurance.
Alex
He aims to deliver efficient and empathetic customer support to users. Alex takes pride in being a reliable problem solver, finding creative solutions to users' queries and ensuring that their needs are met.
User Experience Map
The ideal flow was divided into 3 stages:
-
Pre-onboarding
-
During interaction
-
Post experience
The user thoughts and feelings in each action stage was identified to come up with areas of opportunities.
Based on these opportunities a list was prepared of features and UX recommendations.
UX Recommendations and Features
Scenarios
Based on area of opportunities, features and UX recommendations, 5 scenarios were created.
John recently had surgery and received a medical bill that he thinks should have been covered by his insurance. He contacts the chatbot to inquire about the status of his claim and to understand why it wasn’t approved.
Emily is planning to undergo a medical procedure and wants to know if it’s covered under her insurance plan. She reaches out to the chatbot to get information about her coverage and benefits for the specific procedure.
Alex, a member, recently underwent a medical procedure. He receives a bill for the procedure and is unsure whether it’s covered by his insurance. he also wants to understand the claims process and how to request reimbursement for his out-of-pocket expenses.
Jessica, a member who recently moved to a new city for work. She needs to find a healthcare provider in her network for an upcoming check-up, and she also has questions about her coverage and benefits.
Mark, a member has recently received his insurance ID card in mail. He wants to understand how to use the card, view his profile information, and inquire about his spending accounts.
Ideate
User Flow
Based on the persona we created, the team worked together and identified the route, the most critical tasks that deliver the most value to the users on the product.
Site Map
Site mapping is always important - it gives designers a bird's eye view of the product, it shows how pages are prioritized, linked, and labeled. The team worked together to pan out the sitemap based on the user flows.
Design
First Iteration Samples
We reviewed everything we had created so far - persona, user flows, and sitemap before we started.
The rest of the team was highly involved, too, during this stage. We communicated efficiently with the client throughout, it was a "wireframe > iterate > re-wireframe" type of process.
Together, we went through 3 rounds of iterations by the end of this stage before the wireframes were approved by the client.
Revised Iteration Samples
I followed our companies own Carelon Style Guide. We had to make sure that all UI elements could work with the Carelon system, aesthetically and functionally. Margins, gutters, spacing, fonts & sizes - everything needed to be in line with the components kit.​
Review and Reflections
Empathy-Driven Design:
This project emphasized the importance of designing user-friendly tools tailored to specific needs in the healthcare sector. The chatbot’s ability to simulate natural conversations and provide step-by-step guidance demonstrates how technology can simplify complex processes, such as filing claims or finding providers. Understanding user pain points—like scattered data and the need for personalized assistance—was essential in creating a solution that feels intuitive and approachable.
​
User-Focused Enhancements:
Incorporating features like progress tracking, interactive checklists, and the ability to edit messages ensures that users remain in control of their interactions. Autocomplete suggestions, recommendation chips, and contextual responses provide efficiency, while visual aids like infographics and maps improve accessibility and engagement. Color-coded overviews and easy transitions between steps further simplify navigation, ensuring users aren’t overwhelmed.
​
Balancing Functionality and Experience:
While implementing advanced features like geolocation, provider comparisons, and telehealth integration, it was crucial to maintain clarity and simplicity. Features like one-click dialing, direct booking, and user feedback mechanisms show how small details enhance the overall experience. The project highlights the value of iterative design—incorporating user feedback and refining features to build a system that not only meets needs but exceeds expectations.
