Agartee Technology Inc. Patient Data Management App

Once the core scope was defined, I moved into structuring the product’s information architecture. My goal was to give each role (nurse, doctor, admin) a clear path to the right information with minimal clicks and no unnecessary cognitive load.

User Flow: mapped out step-by-step journeys for the top 3 critical tasks (logging vitals, reviewing patient status, exporting compliance reports).

Wireframe: low-fidelity layouts to define content hierarchy and navigation.

Interactive Prototype (Figma): click-through prototype simulating the entire workflow for remote testing.

Atoms:

The foundational UI elements, such as buttons, input fields, labels, and icons. Each atom is designed to be simple and reusable.

Molecules: Groups of atoms combined to create more functional components, like a labeled input field or a search bar with an icon.

Organisms: Complex, reusable sections of the interface, such as navigation bars, patient overview cards, or multistep forms. Organisms are assembled from various molecules and atoms.

Templates: Page-level layouts that define content structure and user flow but remain flexible for real data.

Pages The finished screens that users interact with, populated by templates and organisms and tailored to specific use cases.

Once foundational wireframes were validated with stakeholders and mapped to real clinical workflows, I developed an interactive Figma prototype simulating the entire nurse experience:

Login and account creation.

Patient search and selection.

Patient management dashboard with data entry forms.

Alongside remote unmoderated shadowing, follow-up sessions ,and close collaboration with the PM, development team, and stakeholders, I conducted task-based usability tests with actual users to rigorously evaluate key design decisions.

Designed realistic, scenario-driven tasks reflecting primary user goals, such as:

Entering patient vital signs accurately.

Recovering a forgotten password.

Reviewing patient history via the overview dashboard.

Recruited representative users (nurses, doctors, admins) reflecting the app’s diverse technical literacy and contextual constraints.

Observed users completing tasks in a controlled, remote setting with screen recording and think-aloud protocols.

Captured quantitative metrics such as:

Task completion time.

Error rates (e.g., incorrect data entry).

User satisfaction ratings.

Collected qualitative feedback through post-task interviews and surveys.

Designed realistic, scenario-driven tasks reflecting primary user goals, such as:

Confusion in Patient Overview Form

Unit measurements missing or unclear.

Labels not descriptive enough.

Lack of actionable error messages.

“Forgot Password” Process Barriers

Users did not recognize the reset email.

Reset links expired too quickly (10 minutes).

Confusion between a reset before login vs. change after login.

Stakeholder Field Expansion Requests

Desire for more detailed patient data for future AI-powered analytics.

Risk of overwhelming staff with non-critical fields.

Dashboard Performance Bottleneck

Initial design loaded all data at once (vitals, meds, long-term trends).

Dev team flagged this as a cause for slow load times during tests.

The launch marked the start—not the end—of continuous improvement. I established a robust feedback loop built on remote usability tests, A/B experiments, and direct user interviews, allowing us to capture authentic insights from frontline staff across multiple facilities without disrupting care.

Accessibility Improvements:

Early remote A/B tests showed that older clinicians struggled with small text and brief UI cues. By increasing font sizes, simplifying language, and introducing on-hover tooltips.

Glove-Friendly Interactions:

Post-launch, remote usability feedback from staff highlighted continued problems with touch responsiveness when wearing gloves. In response, I rapidly prototyped and launched larger, widely-spaced touch targets. Follow-up remote testing demonstrated a 20% rise in task completion rates for glove-wearing users.

Ongoing Remote Experiments:

I regularly launched A/B interface tests and monitored usage analytics to optimize error rates, navigation times, and feature discoverability—all managed remotely to maximize coverage.

Accessibility as a Continuous Standard:

Instituted quarterly remote accessibility and usability audits, ensuring all updates maintain inclusive design standards as the app evolves.

Despite improvements, I conceptualized a hands-free interface to address the limitations of touchscreens in sterile, high-speed environments. I proposed a transition to natural language voice commands for documenting vitals and retrieving patient data, establishing a vision for future development that prioritizes speed and hygiene.

Collaborating cross-functionally to extract predictive insights from patient data, supporting proactive care.

Leading organizational efforts in accessibility and post-launch measurement, with all research activities conducted efficiently and securely in a remote-first model.

Reduced by 35%, per heatmap analysis, highlighting the success of 48px touch targets in minimizing frustration for glove-wearing nurses.

After launching the Agartee "Patient Data Management" app, task completion rates improved to 36%, the error rate dropped by 32%, and user satisfaction increased to 38%, demonstrating the effectiveness of our user-centered design approach.

These performance improvements were instrumental in securing three new nursing home partnerships, expanding from five to eight facilities after launching the app, and driving a 39% increase in annual revenue growth.