Detail Page – f1.rizwanysf.com

Rethinking the Search Experience on Hucu.ai for Healthcare Professionals

Since its launch, the search feature has empowered clinicians to perform around 80,000 searches, significantly improving accessibility and productivity.

January 23, 2025 – 14min read

Introduction

Overview

Hucu.ai is a HIPAA-compliant platform that enables AI-powered, patient-centered communication between patients and healthcare teams. With over 10,000 healthcare professionals and 130,000 active patients, HUCU.ai supports post-acute collaboration to improve care transitions.

Role

As a product designer, I led design initiatives and worked closely with a researcher, product managers, data analysts, developers, and customer success.

Challenges

Clinicians struggled with Hucu.ai's limited search capabilities.

The search function only allowed searching for channel or patient names, making it difficult for on-call staff to manage patient communication and track updates across multiple facilities.
Our challenge was to design and implement a global-level search feature for Hucu.ai. This feature would allow users to search messages in all patient channels, discussion channels, and direct messages (DMs) across multiple facilities.

Discover

Customer Insights

Time-consuming

Without a practical search feature, clinicians had to manually scroll through lengthy conversation threads to find specific patient information or past messages, which is time-consuming and inefficient.

Tracking Conversations Issues

Clinicians faced challenges in tracking the history of patient interactions, leading to inconsistent documentation and potential gaps in the patient’s medical records.

Back and Forth Frustration

Clinicians were often unsure about which facility a patient belonged to, so they had to check each individually. This process was time-consuming and frustrating for clinicians.

If you're curious about the approach, vision, and more, explore
the complete case study, where I've detailed the entire design process.

Discover

The Framework

Before starting any design, we spent time with developers and cross-functional teams to understand the technical limitations, make sense of workflows, and review existing content. This process created early team-wide alignment, sparked great ideas, and built a strong sense of ownership across various teams.

develop

Low Fidelity

My process involved sketching and flow diagrams and then translating these directly into high-fidelity design comps. Moving straight into high-fidelity designs was relatively easy since I worked with many existing design patterns.

develop

Technical Understanding

Performance bottlenecks

We adopted a lazy loading approach, which ensures that clinicians interact with the search results that are most important to them right away. It loads the data (messages, Images) in low-bandwidth

Search Precision

We set the search limit to a minimum of three or more characters for better and more effective search results.

Default filters

To make the search more effective and quick, we discussed selecting the patient channel or facility where patients belong as a default filter when the clinician is trying to search.

develop

Low Fidelity

After receiving feedback from the team, I proceeded with one of the chosen design directions.

Starting on the Same Page

Measurable Business Impact

Before diving into high-fidelity explorations, I aligned customer solutions with business goals to ensure they effectively addressed customer needs and business objectives.

develop

Low Fidelity

After receiving feedback from the team, I proceeded with one of the chosen design directions.

The EXECUTION

High-Fidelity Explorations

Collaboration was key in our agile teams. As a designer, each feature phase involved gathering requirements, consensus, approvals, detailed specs, and handoffs. Prototyping became my go-to tool for getting honest feedback from the team, ensuring stakeholder alignment, and securing leadership approval. I could easily share prototypes as videos and reuse them for usability testing to keep improving.

The release Day

Releasing Features Was Just the Start

On the evening of 26 April, 2024, we began rolling-out this search feature to our customers and we kept a close eye on our customer listening posts. We knew that we received the customer feedback and we fix the things accordingly.
Fortunately, the feedback that we received from our customers was beyond our expectation, customers had difficulty to finding the search, felt disoriented in the app and could not find the search in to the webApp. After many syncs with customers we going deeper and identify the actual root cause of the problem.

Explore the Challenges Post-Launch and How We Solved Them

In my complete case study, I’ve outlined how we tested our assumptions with customers, conducted usability testing, refined designs based on feedback, addressed challenges post- launch and after the feature was released, and managed the handoff to development. It provides a comprehensive view of the process, from ideation to execution.

The Impact

The precision to succeed

/01

Customer searches for messages have increased by 50%

/02

Customers utilizing the search feature to find patient reports have increased by 30%

Reflect

Outcome

Since the launch of the search feature, clinicians have performed an average of 30,000 searches. Thanks to my teammates’ hard work and passion, we improved key metrics and created a solid UX.

Learn more about how we overcome the challenges after release the  feature in the article on medium

BaqarRaza.

Rethinking the Search Experience on Hucu.ai for Healthcare Professionals

Open to interesting challenges and opportunities