Bridging UX & Wellness for Smarter Digital Solutions

Discover how UX, healthcare, and science intersect in my latest projects below

AI-Assisted Stakeholder Information System

Overview

While working in a healthcare-adjacent environment, I observed that our team relied on a fragmented system to store and retrieve important stakeholder information. Customer, patient, and partner contact details lived across spreadsheets, inboxes, and informal notes, making it difficult to quickly identify the correct person or understand the full context of past interactions.

This lack of a centralized system created frequent confusion, duplicate records, and unnecessary back-and-forth between team members. In a healthcare context, where accuracy, timeliness, and trust are critical, these inefficiencies increased the risk of errors and delayed communication.

I initiated this project to explore how a centralized, AI-assisted stakeholder information system could improve workflow efficiency, reduce communication errors, and give teams greater confidence when managing critical relationships.

The Problem

The existing approach to managing stakeholder information was inefficient and error-prone.

Key issues included:

Contact information scattered across multiple tools
Duplicate records created under slightly different names
Limited context around past interactions
Time wasted searching or asking coworkers for basic information
Increased risk of contacting the wrong person or missing important follow-ups

These problems slowed down daily workflows and created avoidable friction in an environment that touches healthcare operations and patient-related processes.

The User

Primary user:
Project coordinators and account managers working in healthcare or healthcare-adjacent organizations who manage communication with clinicians, administrators, laboratories, and external partners.

User needs:

Quickly find the correct stakeholder by name
View accurate, up-to-date contact information
Understand recent interactions at a glance
Avoid duplicate or incomplete records
Act with confidence when reaching out to stakeholders

When users cannot trust the system to provide reliable information, they rely on memory or informal workarounds — increasing stress and the likelihood of mistakes.

Proposed Solution (High-Level)

To address these challenges, I proposed a centralized stakeholder information system designed to serve as a single source of truth for healthcare teams. The system allows users to search for stakeholders by name and instantly view accurate contact details, role information, and relevant interaction history in one place.

The solution emphasizes clarity, accuracy, and efficiency. Each stakeholder has a dedicated profile that consolidates communication notes, key details, and recent activity, reducing the need to search across multiple tools or rely on memory.

To further reduce errors, the system incorporates AI-assisted features such as smart search suggestions, duplicate record detection, and contextual summaries of recent interactions. These features are designed to support human decision-making rather than replace it, helping users work faster while maintaining trust and oversight in a healthcare context.

Design Process

I began by identifying gaps in how stakeholder information was stored and retrieved across existing tools. Because the primary issue was fragmentation and low confidence in data accuracy, I focused first on defining a clear workflow rather than jumping into visual design.

I mapped a primary user journey that included searching for a stakeholder, reviewing results, and accessing a consolidated profile with key context. Using low-fidelity wireframes in Figma, I designed and prototyped the core screens needed to support this flow.

To explore how AI could responsibly support users in a healthcare context, I introduced an AI-assisted summary on the profile screen. This feature was designed to support human decision-making by surfacing relevant context while maintaining transparency and user review.

Design Artifacts: Low-fidelity wireframes and prototype created in Figma (available upon request).