Home

Product design · Enterprise UX

2026

Transforming reporting through a design system migration

I led the migration of BNY’s treasury reporting product to a new design system, challenging legacy constraints to modernize the experience and establish foundational patterns as the first product to launch the system in production.

Role

UX lead

Team

Product, engineering, design systems lead, design systems engineer

Timeline

Q4 2026

Industry

Financial services

Overview

A design system migration is an opportunity to challenge legacy decisions and rethink experiences around user needs and modern technical capabilities.

A design system migration is an opportunity to challenge legacy decisions and rethink experiences around user needs and modern technical capabilities.

A design system migration is an opportunity to challenge legacy decisions and rethink experiences around user needs and modern technical capabilities.

As BNY’s Global Payments and Trade design practice developed a new design system, Reporting was selected for early adoption to support upcoming features.

What I did

Led the redesign of Reporting, challenging legacy decisions and fostering a design-led approach to product development.

Why it mattered

Prioritized user needs over legacy assumptions, challenging decisions shaped by outdated constraints.

Context

What is reporting?

BNY's Reporting product enables enterprise users to generate, schedule, and monitor reports on their financial activity. Three primary pages support distinct workflows: browsing available reports, managing schedules, and reviewing run history.

Who uses reporting?

Reporting supports both BNY teams and enterprise clients, with workflows and available reports shaped by their roles and permissions.

Internal users

BNY teams accessing reports to support client servicing, financial operations, and internal workflows.

Client users

Enterprise treasury teams accessing financial reports to monitor account activity, manage schedules, and support treasury operations.

Release 1

Primary pages: catalog, schedules, and run history

I adapted a single data table component to support three distinct workflows: finding reports, managing schedules, and reviewing run history.

TreasuryEdge Reports catalog screen
TreasuryEdge Reports catalog screen
TreasuryEdge Reports catalog screen

Key improvements

Surfacing primary actions

Brought frequently used actions directly onto the landing page, reducing the need to navigate into individual report details.

Guiding user attention

Visually emphasized items requiring attention, helping users identify where action was needed.

Balancing simplicity and flexibility

Used progressive disclosure to simplify default views while allowing users to customize their experience through column management and filtering.

Release 2

Creating schedules and running on-demand reports

I rethought how users create and run reports, establishing consistent interaction patterns and clearer system feedback across different report types.

Simplifying everyday reporting workflows

Making validation actionable

Kept primary actions enabled, surfacing errors on submission so users could identify missing information rather than guessing why they couldn't proceed.

Keeping feedback in context

Replaced “No data” toasts with inline messaging so users could understand the outcome without losing their place.

Making progress visible

Added a persistent loading toast so users could track report generation instead of wondering whether it was still processing.

Delivering rules instead of screens

Engineering's dynamic report configurator shifted my approach from designing individual screens to defining reusable interaction rules and conditional logic across report types.

Release 3

Viewing and editing schedules

I designed a shared schedule panel that supports two distinct behaviors: quickly reviewing information and making targeted edits.

Viewing schedules

Editing schedules

Editing in context

A section-level inline editing experience lets users update individual sections without leaving the schedule detail panel.

Surfacing relevant information in context

Tailored how schedule information appears across workflows, providing relevant context without requiring users to navigate elsewhere.

Connecting schedules to past activity

Connecting schedules to past activity

Integrated run history into schedule details so users could understand what their schedules had been doing.

Making existing schedules visible

Making existing schedules visible

Surfaced existing schedules during report creation so users could make informed decisions before scheduling another.

Process

From legacy constraints to design-led solutions

Across Reporting, I challenged legacy behaviors rather than treating the migration as a one-to-one translation. Here's one example:

01: Understanding

Using these findings to challenge system constraints through design

I mapped undocumented system behaviors, uncovering where legacy implementation decisions compromised the user experience.

Existing behavior

The system doesn't check whether data exists for the user's selected report parameters before generating the report.

What if?

What if users could recover from a no-data outcome without restarting?

02: Challenging

Using design to challenge legacy constraints

I used AI to prototype more user-centered experiences, making opportunities for system changes tangible and building product buy-in for a better direction.

What if users could recover from a no-data outcome without restarting?

Legacy: No data as a toast

Redesign: No data as an inline alert

03: Aligning

Balancing the ideal experience with delivery constraints

With product buy-in, I collaborated with engineering to define phased improvements when reaching the desired end state required work beyond the release deadline.

Legacy

Temporary feedback

“No data” appears in a toast after the report closes.

Intermediate

Actionable feedback

Add a “Refine report” action to the toast, reopening the report with its inputs preserved.

End state

Feedback in context

Keep the report open and display “No data” inline so users can refine their inputs.

Impact

Setting the standard for broader adoption

Beyond modernizing Reporting, this project helped shape the new design system and demonstrate its value across BNY's Global Payments and Trade organization.

Building organizational visibility

Building organizational visibility

Showcased Reporting to BNY design leadership and at the Global Payments and Trade town hall, helping build support for broader design system adoption.

Shaping the design system

Shaping the design system

Defined foundational interaction patterns and standards that helped shape the evolving design system beyond Reporting.

First to production

First to production

Reporting became the first product to launch the new design system in production.

Next case studyAccount selection

Product Design · Enterprise UX

2026

Standardizing permission-aware account selection at scale

Designing a scalable account selection system that adapts to complex hierarchies, varying permissions, a datasets ranging from 1 to 30,000+ accounts.

Role

UX lead

Team

Design, product, engineering

Timeline

Q3 2026

Industry

Financial services

Context

Project overview

Skip to solution

BNY’s Reporting product allows enterprise users to run reports on financial activity such as payments and balances. Across these reports, users select which accounts they want to include. I designed a new account selection experience to support increasingly complex account structures and selection needs.

What I did

Designed a new account selection component for BNY’s Reporting product that later scaled across the online banking platform.

Why it mattered

Existing selection patterns did not support the scale and complexity of BNY's new virtual account offering.

Account structure and identification

Accounts can be physical or virtual, with virtual accounts nested under physical accounts. What users see and how accounts are grouped depends on their permissions.

Users rely on name, number, currency, type, and hierarchical relationships for accurate identification.

Permission-based user types

Instead of fixed personas, users are defined by role and access.

Internal users

BNY teams supporting operational and servicing workflows

Client users

Enterprise treasury teams managing accounts and reporting workflows

External users

Third-party users (such as contractors) with limited or controlled acces

Problem

User need

Quickly identify and select relevant accounts across hierarchical structures

Business need

Support consistent account selection across growing datasets and workflows

I had two key challenges to solve

Challenge 1

Account identification

How might we uncover and surface the key attributes users rely on to quickly and accurately identify their accounts?

Challenge 2

Account selection

How might we adapt this experience to different dataset sizes and selection intents, ranging from individual accounts to large sets?

Understanding user behavior

I spoke with product owners and analyzed existing workflows to better understand user behavior.

Insight

Account selection is a retrieval and validation problem, not a discovery problem.

Users often arrive with externally maintained account lists, known account numbers, and specific accounts they intend to select.

Solution

Three hierarchy patterns for different levels of access

Users see different account structures based on their permissions. I defined three hierarchy patterns to account for these variations: grouped, semi-grouped, and flat.

Hover over each view to see which accounts that user can access

What does this look like in a UI?

Supporting different selection intents

Dataset size and the task at hand shape how users approach account selection, requiring different ways to find and select accounts.

Paste-to-match selection

Users with large, externally maintained account lists can paste them in to automatically match and select accounts.

Attribute-based filtering

Users can filter by common attributes to quickly select groups of relevant accounts.

Hierarchy-aware search

Searching for a physical account returns its child accounts, allowing users to select the entire group at once.

Process

As the problem grew, so did the solution

What began as a lightweight enhancement to an existing account dropdown evolved as new requirements surfaced, ultimately requiring a new, scalable selection model.

Version 0 (legacy)

Existing dropdown

Challenge

Built for small account sets and physical accounts only.

Approach

Simple dropdown with account numbers only.

Version 1

Adding account identification

Challenge

Users needed additional identifiers to distinguish accounts.

Approach

Added account name and currency to improve recognition.

Version 2

Scaling beyond dropdowns

Challenge

Increasing account volume and virtual accounts made simple dropdowns difficult to use.

Approach

Introduced searchable table patterns, larger selection areas, and selected/all states.

Version 3

Supporting hierarchical accounts

Challenge

Users needed visibility into relationships between physical and virtual accounts.

Approach

Added a detail view with supporting information.

Version 4 (final)

Supporting large account sets

Challenge

Manual selection became inefficient at enterprise scale.

Response

Added paste-and-match bulk selection workflows.

Curveball!

I had to switch design systems!

Version 0 (legacy)

Existing Dropdown

Challenge

Built for small account sets and physical accounts only.

Response

Simple dropdown with account numbers only.

Version 1

Adding Account Identification

Challenge

Users needed additional identifiers to distinguish accounts.

Approach

Added account name and currency to improve recognition.

Version 2

Scaling Beyond Dropdowns

Challenge

Increasing account volume and virtual accounts made simple dropdowns difficult to use.

Approach

Introduced searchable table patterns, larger selection areas, and selected/all states.

Version 3

Supporting Hierarchical Accounts

Challenge

Users needed visibility into relationships between physical and virtual accounts.

Approach

Introduced grouped, semi-grouped, and flat hierarchy patterns.

Version 4 (final)

Supporting Large Account Sets

Challenge

Manual selection became inefficient at enterprise scale.

Approach

Added paste-and-match bulk selection workflows.

Future enhancements

Things I'd like to explore further:

Do users need a separate selected view?

Can selections can be surfaced effectively within search and filtering, or is a dedicated review step better for supporting reviewing selections?

Is there a use case for saving and reusing account groups?

If users maintain external lists of accounts to report on, could we allow them to create and store reusable groups within our product to support Reporting and other areas of online banking?

Takeaways

Key Learnings

AI assisted problem solving

Leveraging AI to clarify requirements, explore solutions, and rapidly prototype ideas to drive stakeholder alignment.

Navigating ambiguity

Building on iterations with intention, adapting the solution as new requirements emerged and the problem became clearer.

Systems thinking

Turning a single-use solution into a scalable component that works across products and workflows.

Balancing use cases

Prioritizing common use cases while designing scalable, flexible solutions that gracefully accommodate edge cases.

Previous case studyReporting migration
Next case studyConductor

Like what you see?
Let’s connect.

© Rebecca Skier 2026

Created with love, Framer, Claude, and lots of caffeine.

Like what you see? Let’s connect.

© Rebecca Skier 2026

Created with love, Framer, Claude, and lots of caffeine.

Like what you see? Let’s connect.

© Rebecca Skier 2026