ifatarg@gmail.com

Home

BRAND

RESPONSIBILITIES

Lead Product Designer
User research
Design system
Prototyping
Usability testing
UI/UX QA before production

CREDITS

PM Maya Cohen / Ollie Berniker
Teach Lead Avitan Barnes

Launching a New Analytics Experience-Driving 150% growth in user engagement

Making a highly complex system feel flexible and approachable, even to users unfamiliar with similar platforms

Overview

I led this project from the ground up, designing an advanced analytics module embedded within a broader healthcare platform.

Analytics App is a tool that empowers clinicians and data users to create, explore, and share dynamic dashboards and clinical data visualizations, blending familiar paradigms from commercial analytics platforms with domain-specific functionality tailored to medical needs.

+150%

In its second year post-launch, the platform experienced a 150% increase in user engagement

Callenges

Design an intuitive, scalable platform tailored to medical workflows, allowing users to build, customize, and share dashboards with minimal data preparation, while supporting advanced features that balance flexibility and ease of use.
Enabling users of varying skill levels to explore and act on complex clinical data.

User Flowchart

User Flowchart

User Research & Insights

Before design, I interviewed clinical users and data professionals to understand their pain points.

Key insights included long data prep times (1–4 hours), challenges with live updates, data privacy, and inconsistent access.
62% showed interest in AI tools, and 66% wanted tailored dashboards to streamline workflows.

Following two rounds of user testing and two rounds of A/B testing – with iterative improvements throughout – the final version reached an 87% success rate and was released to production.

User Testing results

Design solutions

Turning Confusion into Clarity

The filter and comparison feature posed usability challenges during testing.

In early iterations, users struggled to grasp how to filter original data and compare it to a modified set – only 71% understood the workflow in Iteration 1, and just 50% in Iteration 2 due to confusing tab structures.

In Iteration 3, I redesigned the flow as a guided workflow, requiring users to apply a filter before enabling comparisons. This streamlined approach improved comprehension to 88%, with users successfully completing the task and understanding the connection between filters and comparisons.

Prioritizing Usability Across Experience Levels

To balance the needs of both experienced and less technical users, I surfaced the core functionality upfront, ensuring quick access and ease of use.

More advanced, complex features were placed within a collapsible section, keeping the interface clean and approachable for new users, while still providing the depth and control that power expert users.
This approach reduced cognitive load without sacrificing functionality.

AI Chart Generator

A new feature that allows users to describe the chart they want in natural language, and the system builds it automatically.
This helps non-technical users access complex visualizations without understanding query syntax.

Dashboard Duplication

Users can duplicate existing dashboards or individual charts to streamline exploration

Sharing is caring

Users can have full control over dashboard visibility – they can choose to share dashboards with their CareTeams or keep them private, ensuring complete ownership and transparency over who can access their data.

Dashboard grid configurations

Users can configure grid layouts, allowing tailored arrangements for different workflows.

Data Source Aggregation & Splitting

Users can aggregate data from multiple sources
(e.g., cross-facility comparisons) or split data views by team membership (e.g., different CareTeams).
This enables granular insight into both high-level trends and team-level variations.

Outcome Highlights

  • In its second year post-launch, the platform experienced a 150% increase in user engagement, indicating deeper, more frequent usage across the board.

  • Early adopters appreciated the clinical specificity of the tool compared to commercial data tools.

  • The platform eliminates the need to jump between tools and reduces the time required to produce publication-quality clinical charts.

  • Sharing logic and team-based permissions were seen as critical enablers for collaboration.

  • The AI Chart Generator enabled users to shift their focus from manual tasks to efficient data exploration
 

Reflection

Designing this analytics tool was a deep exercise in balancing complexity with usability.

The biggest success was making a highly complex system feel flexible and approachable, even to users unfamiliar with analytics platforms.
By grounding the experience in real-world medical workflows and listening closely through every test cycle, we created a solution that not only works, but works for clinicians.