One system to bring order to health insurance ops
VizCare
What's missing? One reliable place to work
Health insurance workflows were spread across separate tools. Agents were wasting time switching between them and fixing errors. I built Vizcare with a team to connect everything in one place, allowing operation teams to focus on supporting people, not troubleshooting systems.
The Team
Me (Lead Product Designer)
Sanaa Ayesha(Junior Product Designer)
Nishant Shrivastava (Project Manager)
Sana Ru (Lead Frontend Developer)
Shardha Sapra (Senior Frontend Developer)
Timeline
2023 June - December
Background
US health insurance operations rely on disconnected, aging systems
Most insurance workflows run on tools built at different times by different teams.
They're rarely designed to talk to each other, slowing tasks, spreading errors, and making routine work harder than it should be.
Problems
What slowed agents down wasn't complexity. It was a disconnection
Each workflow, like member data, commissions, claims, and sync, ran on a separate system. Because nothing was connected, even basic tasks required switching tabs, double-checking info, and fixing avoidable errors.
Split Systems
Benefits, commissions, and claims each lived in their tools, making it hard to even start a task without switching between them
Manual Process
Even simple status checks meant bouncing between systems. There was no seamless path, copy, paste, and recheck
No Unified View
Agents had no single place to view synced data, coverage info, and claim status, which often led to critical details slipping through the cracks
Goals
Internal tools didn't connect, so we set out to make key operations easier to follow, edit, and rely on
01
See Status at Glance
Let teams instantly spot what's changed and what's pending, no more tool-hopping to check status
02
Edit without Losing Context
Allow updates while viewing the right case for users not to lose track of switching screens
03
Share a Single Source
Give teams one up-to-date view to work from, avoiding outdated info and scattered versions
System Snapshot
Tracing where user-facing errors began across disconnected tools
User Research
Understanding team-specific blind spots through interviews and tool audits
I wasn't designing for just one "user." Each team touched the data at different points with different goals. Mapping the full picture required talking to them in their context with 10 team member interviews (2-3 per role), tracing 3 recurring sync issues traced across 6 recorded support calls, and reviewing the internal tool usage log with the product and engineering teams.
Interview synthesis & feature prioritization
Each feature stems from specific behavior patterns observed across teams
Cross-team interviews and usage logs surfaced key gaps. Each row maps a pattern to a focused solution.
Key Insights
Hidden Breaks
Teams acted on outdated or partial info without realizing anything was wrong
Fragmented Views
No single team had visibility across membership, claims, and support; misalignment was inevitable
Late Fixes, High Cost
By the time issues were caught, they had already disrupted calls or payments
Features Solutions
Policy Trace View
Help teams pinpoint sync failures across tools and resolve them without chasing data across different screens
Unified Edit Layer
Centralized place to view and update sync logic across tools, keeping teams aligned and reducing miscommunication
Call-Ready View
Gives agents a clear summary of sync status, delays, and flags to help agents resolve issues on the spot and avoid escalation
Information Architecture
Mapping the system backbone for a shared vision
Before diving into design, I worked closely with business analysts, engineers, and product managers to define how each platform serves distinct user groups and operations. By aligning platform responsibilities and cross-functional logic early on, we ensured consistency and avoided fragmented workflows that often emerge from tool-by-tool thinking.
Systematic UI Decisions
Structuring UI to bridge data gaps
I turned logic gaps into targeted screens, showing each UI element resolved a specific problem.
Making payout and coverage logic visible at a glance
Enrollment agents often wasted time flipping through tabs or opening external files to confirm basic benefit or payout rules. I combined plan, eligibility, and payout info into one view for faster checks. Agents can scan plan and payout details in one place without switching between tabs.
Membership & Coverage
I split coverage and commission into focused views so agents can check both faster, no more tool switching.
Membership View
For MemberX / BrokerX (Xperience),
Logic by Empower
Commission View
For Empower agents,
Track payouts & rules
Centralizing sync errors across tools
I replaced three separate sync views with a dashboard that connects volume summaries with file-level and process-layer errors. This decision enables agents in tracing issues more quickly without switching between tabs or tools.
Unified Sync & Error Dashboard
From volume overview to individual record trace, all sync layers are available in one structured view.
Transactions Overview
For Empower agents (Connect),
Error summary by sync source
Sync Dashboard Summary
For Empower agents (Connect),
Volume-level sync breakdown by mapping
Routing Flow View
For Empower agents (Connect),
End-to-end views of sync routing logic
Giving agents the whole picture during support calls
Agents used to toggle between tabs to find details during the call. To fix this, I unified context, like member info, contact history, and AI replies, into one screen to reduce delays and mistakes.
Context-Rich Contact Center
This screen is a centralized view of member details, AI tools, contact history, and message tracking, designed to minimize back-and-forth and improve call handling.
Contact Center
For contact center agents (Empower),
One unified view of member, notes, and live actions
Contextual AI Tools for Agents
Agents can now:
"Reply with AI" button opens contextual suggestion mid-call
Edit tone or wording instantly with AI-assist tools
Avoids switching apps, write and review all in one screen
Duplicate Question Detection
Detects repeated questions during the call and instantly surfaces past answers. It helps reduce backtracking and resolve calls faster
Structuring enrollment into clear, guided steps
Users struggled with scattered, unstructured forms that made it unclear which fields were required or what came next. I redesigned the process as a step-based flow, prompting only one decision per screen to prevent errors, missed inputs, or hesitation.
*Note: This screen was introduced later in the project to support employee-initiated enrollment.
Step-Based Enrollment
The flow separates decisions like life events, coverage, and dependents into distinct stages to guide users with minimal friction.
Life Event Intake
For EmployerX (Xperience),
Logic by Empower
Step 1 starts by capturing intent, prompting users for the life event reason, dates, and docs needed to proceed, while signaling what's next
Outcome
Real gains and minor blind spots
I ran an unmoderated usability test with 10 internal users from the research process, using anonymized patient data and real logic from existing policy files. Most users navigated key workflows like Commission and Sync, smoothly without extra guidance.
62%
Faster error resolution
48%
Less data re-entry
1.3 x
AI-assist used more frequently post-launch
4.6 ★
Avg clarity rate
However, two blind spots surfaced:
A few users hesitated at entry, unsure where to begin
The AI assistant handled 1:1 flows well, but felt too generic for edge cases
Next Steps
While internal users navigated the tools with ease, a few friction points remain. Next, we plan to:
Onboarding Direction
Help users begin with clear first steps
Smarter AI Responses
Train AI with real cases to better handle edge scenarios
External Pilot Test
Test usability with external users