Zemplee: AI Health Tech

Project Overview

Zemplee is an AI-powered health tech startup that enables aging in place through remote patient monitoring. Their platform helps care coordinators and nurses track patient health data, manage alerts, and coordinate care plans.

I was brought on as the sole UX Researcher to tackle a critical efficiency problem: the platform's alert management system was consuming too much of nurses' time, leaving less time for actual patient care.

My Role

UX Researcher: Product Research, Secondary Research, Stakeholder Interviews, User Interviews, Contextual Inquiry, Accessibility Audit, Usability Testing

Team

Sole UXR collaborating closely with Product Manager and Engineering team

* Due to NDA restrictions, this case study provides a high-level overview of my research approach and impact. Detailed methodologies, deliverables, and product visuals remain confidential.

Neurodivergence, encompassing conditions like ADHD and autism, affects over 8.7 million adults in the US. Due to the pandemic, there has been a rapid shift to remote work which is now embraced by over 33% of the workforce. In this shift, Slack has become a widely used workplace platform; however, our initial research indicates Slack does not address the diverse needs and challenges of neurodivergent individuals, preventing a supportive and productive environment for all users.

Key Research Impact:

53% reduction in clicks per alert (17 to 8 clicks)
45% faster alert resolution
2+ hours returned to nurses for patient care daily
120+ accessibility issues identified with product-wide accessibility standards and updated brandbook created

The Challenge

Care coordinators were spending 15-20 clicks to manage a single patient alert, with each patient generating multiple alerts daily. The problem intensified on Monday mornings when weekend alerts accumulated, creating hours of backlog.

My task: Reduce the clicks per alert by 50% while maintaining the EHR-like (Electronic Health Record) experience nurses expect and improving overall platform usability. While accessibility wasn't part of my original scope, I proposed conducting a comprehensive accessibility audit after recognizing the opportunity to improve the platform for all users, not just those with specific accessibility needs.

Phase 1: Understanding Current State

Reasoning: I started with secondary research to understand the current platform and user experience, identifying what had already been explored and where opportunities for new learning existed.

Methods:

Secondary Research & Synthesis

Analyzed 4 existing user interviews and 3 surveys from previous quarters
Identified recurring themes around "fighting the system" and workflow efficiences
Discovered care coordinators (nurses, clinicians) manage 150-175 patients each and feel overwhelm due to time allotment vs. patient volume

Literature Review & Competitive Analysis

Conducted in-depth literature review on emerging and existing technologies in aging-in-place and health monitoring care
Analyzed competitive remote monitoring platforms and EHR systems, noting key features and opportunities
Identified best practices in healthcare alert management systems
Synthesized findings into design principles for healthcare interfaces

Research Artifact Refinement

Updated existing personas to reflect actual behavioral patterns discovered in interviews
Redesigned existing survey instrument to include standardized scales for quantiative analysis and workflow comparison questions to measure concrete efficiency gains and system use impact
Incorporated technology comfort assessment to understand initial user onboarding friction and segment user responses

Stakeholder Alignment

Conducted 2 interviews with CEO and board members to understand key business and growth goals
Aligned on business goals of scaling patient numbers and care functionalities
Aligned on success metrics of efficiency gains and user satisfaction

Key Insight: This phase of research revealed a pattern of nurses feeling overwhelmed by patient volume versus time, but no one had quantified exactly where that time was being lost. This gap directed me toward observational research in Phase 2.

Phase 2: Identifying Barriers

Reasoning: In this stage, I focused on direct observation to unearth what our users can't, or don't know to, share during interviews. My goal was to see the platform through the nurse's eyes in their actual workflows, understanding not just what was difficult, but current workarounds they created.

Methods:

Accessibility Audit (Self-Initiated): While I was familiarizing myself with the platform, I recognized key accessibility issues. I took the initiative to propose an additional project (which was approved by the CEO herself!) to improve the platform through an accessibility audit

Created a project proposal and presented to stakeholders, showing value of accessibility improvements that impact all users
Identified 120+ accessibility issues through WCAG audit, collaborating directly with developers to resolve these issues
Discovered visual navigation and visibility issues in key work flows
Found inconsistency in branding materials

Contextual Inquiry: I conducted 3 remote observation sessions with nurses during actual workflows

Discovered key workflow barrier where alerts pile up over weekend creating a near-impossible Monday morning flow for nurses
Found key workflow barrier of nurses opening and switching between tabs rather than using within-tab navigation caused by lack of discoverability. Nurses were not aware of an existing functionality creating added time and blocks.
Mapped the complete 17-click alert resolution journey
Observed nurses' workarounds to system like physical sticky notes on monitors or adding notes to open text sections
Identified confusion around functionality and terminology of new features that caused disuse

Key Insight: Nurses weren't avoiding features because they didn't want them. They actually didn't know they existed, let alone how to user them. This discovery shifted our focus from adding new features to making existing ones discoverable. Additionally, this phase revealed how nurses worked around the platform rather than with it, running into time barriers like clicking between tabs.

Phase 3: Design & Validation

Reasoning: Here, I moved from observation to action, translating insights into tangible design improvements. My goal was to validate that our redesigned workflow would actually reduce cognitive load during high-stress Monday mornings. I needed to ensure the solutions respected nurses' existing mental models while eliminating the friction points we'd identified.

Methods:

Workflow Redesign: I collaborated closely with my PM to redesign the existing workflow and create a new user journey mapping

Consolidated the 17-click journey into an 8-click streamlined flow
Unified alert viewing and resolution into a single interface, eliminating tab-switching
Implemented bulk action capabilities for managing multiple similar alerts
Integrated "sticky note" functionality based on observed nurse behaviors
Maintained familiar EHR patterns while removing redundant navigation steps

Collaborative Implementation

Created tickets and implementation guidelines for development team
Proposed new brand colors and brandbook meeting WCAG AA standards (maintaining the purple loved by nurses while fixing contrast issues)
Worked directly with developers to prioritize fixes based on user impact

Usability Testing: I tested the redesigned workflow with 4 nurses to validate the three key hypotheses below and understand how this redesign was received by the nurses.

Guiding Questions:

Efficiency gains: Would the new flow actually reduce time-on-task during Monday morning rush?
Mental model alignment: Did the unified interface match nurses' expectations?
Feature adoption: Would nurses use the new sticky note feature over physical notes?

Testing Approach:

Created realistic scenarios based on Monday morning alert backlogs
Measured task completion rates (improved from 75% to 100%)
Tracked time-on-task (3.2 minutes to 1.7 minutes per alert)
Gathered qualitative feedback on unified alert view and color system

Key finding: "I wish there was one page where I can see and resolve all the alerts" from the first interviews became "This is exactly what we needed"!

Research Impact

My research directly transformed how nurses interact with the platform, creating measurable improvements in both efficiency and accessibility. I saw an excitement in the nurses for the workflow improvement, directly tied to the increase of usability and workflow efficiency.

The work established foundational research processes and accessibility standards that continue to guide product decisions at Zemplee past the duration of my internship. Here are some highlights:

Quantitative Results:

53% reduction in clicks (from 17 to 8 per alert)
45% reduction in time spent per alert
120+ accessibility issues identified and prioritized
2+ hours returned to patient care per nurse daily

Qualitative Results

Improved nurse satisfaction with platform ("We want to be the best we can be for our patients. This is exactly what we needed")
Established product-wide accessibility standards and created an accessibility audit guidelines for future work
Created and optimized new and existing research materials such as surveys, interviews, user flow diagrams with training documentation
Updated brandbook guidelines meeting accessibility standards

Reflection

I am lucky and grateful to have been given so much trust and ownership over my research during my time at Zemplee. Researching in this healthcare space that positively impacts so many users and being able to focus on improving accessibility and usability is exactly what drives me as a researcher. I learned and grew so much during this internship through mentorship under my PM, Ishanya, and through working to improve the experience of nurses who do such important work.

Here are my key takeaways:

Accessibility is Universal Usability: The accessibility issues on the platform weren't just edge cases. They were issues that can affect every user, especially during stressful morning rounds with older technology and poor lighting. Focusing on improving accessibility directly improved the usability and efficiency of the platform.
Observing before Optimizing: Nurses developed workarounds for good reasons. Understanding why they used sticky notes rather than an in-built feature or why they struggled with navigation led to better feature design instead of simply "fixing" the interface.
Design for the Worst Case: Monday Mornings were a clear problem area for our nurses that stood out through every conversation. By optimizing for Monday morning alert backlogs, we created a solution that performed well for all the other days of the week as well.