Dashboard Design

Dashboard Design

Streamlining Compliance Monitoring: A Nursing Home Oversight Dashboard

Role

Lead UX/UI Designer (solo)

Tools

Tableau, Figma

Timeline

8 Months

Impact

80% faster data access · WCAG-compliant · Adopted by 4+ teams

The problem

In the aftermath of the COVID-19 crisis, NYS needed tighter oversight of long-term care facilities. But enforcement and analysts were working with:

Outdated Excel sheets with no version control
No centralized source of truth for facility data
A fragmented Tableau file with randomly placed tables and maps
Analysts struggled to locate high risk facilities quickly
Visual inconsistencies and non-compliant accessibility
Data interpretation required excessive manual filtering

This lack of clarity delayed reporting and decision making, during a time when every day mattered for vulnerable nursing home residents.

Users

The primary users were analysts and supervisors within the Office of Aging & Long Term Care. Their workflow was investigative, not passive. They needed to:

Move quickly between statewide views and facility-level detail
Asses risk at a glance
Trust that the data they were acting on was current

Key Insights

Research consisted of stakeholder interviews with analysts and supervisors and a content and accessibility audit.

Critical data was scattered across multiple Excel spreadsheets with unclear ownership and unreliable update schedules
Analysts had no way to assess data currency without hunting for it manually
The existing layout did not reflect how analysts actually scanned for risk
The interface had accessibility gaps and no consistent application of DOH branding

Key Design Decisions

Decision 1

Centralizing data with clear hierarchy

The paper process required two separate forms: a standard registration and a multisite supplement. Many applicants didn't know the second form existed; others filled it out when they didn't qualify. In the digital application, the multisite section appears only when eligibility criteria are met. Eligibility logic moves from the applicant to the system.

Decision 3

Distinguishing fixed facts from negotiable decisions

The agency also wanted to collect test procedure information before secondary sites, with an upfront error if an applicant exceeded 15 tests. I pushed back. Ownership type is fixed: you cannot change your organization type to unlock multisite. Test count is a live decision: an applicant with 16 tests who wants multiple sites can choose to reduce to 15. An upfront warning gets dismissed. A forced choice gets made.

Decision 2

Sequencing to surface blockers early

The agency wanted lab information collected first, then director and ownership details. The problem: an applicant could complete the entire form before discovering they don't qualify. My position was to surface disqualifying information as early as possible. Hit the wall at question three, not question thirty. The agency agreed. Director information now appears before lab details; ownership type appears before the secondary site section.

Decision 4

Replacing an infeasible input with a purpose-built component

With 30+ test procedures, a standard checkbox list was unworkable. Working with developers, a solution was created, a bucket component: available tests on the left, selected tests on the right. The one place the NYBE design system needed extending. Knowing when to push past platform defaults, and having the relationships to make it happen, was what made it feasible.

Outcome

Required field validation, conditional eligibility logic, and mandatory document uploads ensure a reviewer can no longer receive an application with missing information. The applicant experience shifts from "fill out what you know and mail it" to a guided flow that enforces completeness at every step, a fundamental change in how the form enforces compliance.

Reflection

The most valuable skill on this project has been knowing when and how to push back on stakeholders. The agency's instincts consistently reflected how they process applications internally, not how an applicant moves through a form. Reframing around user behavior was more persuasive than citing design principles.

Designing with system consistency in mind has also mattered. With 14 applications across 4 permit types, decisions made here carried forward across the entire system. The amendment application, consolidating 6 paper forms and 3 undocumented processes into a single digital flow, was the most complex piece of this work and will likely warrant its own case study.