Overview
One of the world's largest pharmacy chains invited us to develop initial MVP designs for their next-generation Call Center Agent Portal. The goal was to unify multiple outdated systems and workflows into a single source of truth. This initiative aimed to streamline operations and reduce the learning curve by creating a simplified, intuitive system with fewer clicks. The client was also eager to explore AI integration to enhance employee productivity wherever possible.
As the sole designer on the team, I was responsible for the entire design workstream. I collaborated closely with a Product Manager, a Senior Strategist, and a Junior Strategist to manage client relations, define the product roadmap, facilitate workshops, and craft UI/UX designs.
Challenges
Navigating an Unfamiliar Business Process
Pharmacy call center agents often used industry-specific terminology. Moreover, the client's system combined off-the-shelf IVR software with custom-built pages that had limited documentation. Given my lack of prior experience in pharmacy or IVR systems, understanding the business process well enough to make impactful design enhancements posed a challenge.
Compounding this, I entered the project with just one month remaining to replace a designer who struggled to capture the client's vision for the product. As a result, there was considerable scrutiny on the rationale behind my design choices.
To resolve these challenges, I drew on my past experience with enterprise software and looked for commonalities. For instance, I found many parallels between the process of managing prescription refills and workflows for inventory management and purchase orders that I've previously worked on. I also saw many similarities between IVR software and widely used web meeting platforms like Google Meet or Microsoft Teams. This enabled me to collaborate with internal subject matter experts to brainstorm ideas and adapt design patterns from these comparable products to meet the specific needs of the client, thereby increasing their trust in the designs.
Current UI lacks visual hierarchy and has an overwhelming number of buttons
A screen I designed before for managing inventory purchase orders shares many similarities in structure and functionality
Redesigned screen focuses only on essential information and actions for this step in the flow, with improved visual grouping and additional features like search and sort to help users quickly find desired information
Call and chat panel designs inspired by Microsoft and Google products
Finding the Right Balance for AI
Another challenge we encountered was determining the appropriate areas to integrate AI. While stakeholders were eager to leverage AI advancements, they lacked a clear vision of its specific applications within the portal. The initial inclination was to incorporate AI extensively to potentially replace labor and maximize cost savings. However, stakeholders often underestimated the costs associated with implementing and maintaining AI, as well as its limitations.
Our team proposed a user-centered approach based on specific pain points in the current user experience. We aimed to identify areas where relatively straightforward AI implementations could enhance worker efficiency, rather than fully automating tasks traditionally performed by employees.
For example, we observed that agents often kept multiple open notepads on a second screen filled with canned chat responses. Further investigation revealed that specific guidelines for agent responses were stored in Word files on an online drive that was challenging to search. To save time, agents manually copied commonly used responses onto notepads for quick access. However, this workaround posed a risk of using outdated responses because agents stopped checking the original documentation file for updates.
After consulting with technical experts to assess feasibility, we decided to implement two AI-driven features in our design to address this issue. The first feature provided contextual, AI-generated suggested responses directly within the chat window, allowing agents to skip manual copying and pasting while ensuring the information is up-to-date. The second feature introduced a Knowledge Article Panel where AI dynamically selects relevant documentation materials by analyzing the live conversation transcript.
An agent's desktop cluttered with many notepads containing canned chat responses
AI-generated suggested responses based on chat context and guidelines documentation
Dynamic Knowledge Articles Panel where AI surfaces relevant documents in real-time based on analysis of call/chat transcripts
Outcome
We successfully completed over 20 MVP screen designs, covering workflows for the dashboard, customer profile, and call/chat functionalities. The client stakeholders were very pleased with the outcomes, noting that the designs accurately reflected their vision. They were especially grateful for our guidance on AI, which greatly helped them make informed decisions. Looking ahead, the client stakeholders are excited to showcase the MVP concepts to top-level leadership in pursuit of securing additional funding.
Praise from PwC Partner overseeing the project
Prototype showing the prescription refill workflow
Dashboard mockups
Customer profile & create prescription mockups
Call panel mockups
Chat panel mockups
