Josh Koh
All Work

Papakilo Database · Office of Hawaiian Affairs

Making Hawaiian History Easier to Find

Project Tags:UX ResearchInfo ArchitectureGovernment UX
Papakilo Database / Cover4320 × 1869
The redesigned Papakilo Database on a laptop and two phones.

01 Overview

A Database Unchanged Since 2009

2009the year Papakilo launched, and it had run essentially unchanged since
5people interviewed in depth: 3 internal staff and 2 external users
4 → 1separate search entry points, consolidated into one search bar
~5minto find a document in testing, down from 20–25 minutes

02 The Stakes

“What Is UX?”

Papakilo is the database Native Hawaiians use to access genealogical records, historical land documents, and cultural knowledge connecting them to their heritage. It had run essentially unchanged since its 2009 launch.

There was no UX function at OHA. When I proposed exploring a redesign as part of my UX certification, my team’s first question was, “What is UX?”

The real assignment

This wasn’t just a design problem. It was a chance to introduce user-centered thinking to a government agency that needed it and had no language for it yet.

An annotated UX audit of the Papakilo homepage and filtered search pages, covered in sticky-note callouts.
The first UX audit: key issue areas across the homepage and filtered search pages.

03 Research

If the Expert Struggles, Everyone Does

I interviewed 5 people. Archival research is a niche enough area that I chose depth over breadth.

Internal staff · 3

  • Luci MeyerCultural archivist, Papakilo’s daily expert user
  • Staff memberFellow multimedia designer
  • Staff memberBoard of Trustees staff

External users · 2

  • College studentResearching Hawaiian music history
  • First-timerNo archival research background at all

If the expert user struggles, casual users don’t stand a chance. Luci put it directly.

“There’s no way to help people keep track of information they want to keep referring to. They have to restart the search process every single time.”

Luci Meyer, Cultural Archivist

“Why are there four different ways to search? Which one gives me the best results?”

College student, external user

The external user hit the same wall from a different angle, unsure which of four separate places to even begin a search.

Illustrated portraits of the internal staff and external users I interviewed.
Who I talked to, and what each person needed from Papakilo.

04 The Real Problem

Not a Visual Problem. An Architecture Problem.

It wasn’t a visual design problem. It was an information architecture problem hiding inside a search problem. Fragmented entry points, no memory of a user’s search context, and filter categories that used internal jargon nobody outside the archive team understood.

I proposed consolidating four separate entry points into one search bar. I replaced the filtering system, which let users search only by collection, every single one listed out in a single dropdown, with four categories organized around how people actually look for things.

Search, restructured

4 places to start. Pick one before you know what you’re looking for.

  • Search by collection
  • Search by location
  • Search by newspaper
  • Advanced search

Filter by collection

…every collection, one long list, in archive-team terms

Low-fidelity user flow sketches exploring a single search entry point.
Hover the sketches to look closer. Early user-flow sketches for getting into search from one place.

05 The Proposal

Three Changes, Each Traced to a User

With the structure settled, I redesigned the screens around it. As the sole designer on this internal initiative, I owned the complete visual system and every design decision. I focused on three core improvements, and each one answers something a real person told me.

SOLUTION 01

One Search Bar Instead of Four Front Doors

Four competing search methods created decision paralysis. Nobody could tell which path would give them the best results.

The Old Design

Four different entry points (by collection, by location, by newspaper, advanced search) forced people to choose a search method before they knew what they were looking for.

The Proposed Design

One centralized search bar with progressive filtering. Start with a simple query, then refine as needed, so people can begin searching right away.

After: the proposed homepage with one central search bar.Before: the old Papakilo homepage with four separate search paths.OldProposed

Who this answersThe college student who asked, “Why are there four different ways to search?”

SOLUTION 02

Searches That Remember Where You Left Off

Even a power user like Luci had to restart her searches every time. There were no saved searches, no history, and no way to build on previous work.

The Old Design

Every search started from scratch. No memory of what someone had searched for, no saved parameters, no personalization.

The Proposed Design

User accounts with saved search parameters, search history, and recommendations based on the collections someone uses. Onboarding asks about research interests to surface relevant content up front.

After: proposed mobile screens with saved searches and history.Before: the old mobile search screens with no saved searches.OldProposed

Who this answersLuci Meyer, who watched people “restart the search process every single time.”

SOLUTION 03

Partners and Updates You Can Actually Scan

Dense blocks of partner information and poor visual separation made the interface overwhelming.

The Old Design

Database partners displayed as undifferentiated text blocks, with no clear way to scan for a specific organization or understand what each partner offered.

The Proposed Design

Partners shown as expandable cards organized by category, with clear hierarchy and breathing room between sections. “Latest updates” got more prominence so it’s easy to scan.

After: proposed partner cards organized by category.Before: the old partner page as dense blocks of text.OldProposed

Who this answersThe first-time user who said, “I don’t know where to start, so many things competing for attention.”

The Papakilo style guide: colors, type, buttons and components.
The style guide behind the proposal.

06 Testing + Refining

Four Testers, Five Minutes, No Help Needed

To check that the redesign worked and felt intuitive, I ran usability tests with 4 new participants, none of them from the original research group.

The task: find a specific historical document using the redesigned search and filtering.

Testing also surfaced three things to refine before I presented the work to stakeholders.

Before and after: the mobile menu with clearer labels and grouping.

REFINEMENT 01

Clearer Menu Pathways

Some testers weren’t sure how to reach certain features, so I improved the labeling and organization of the navigation structure.

07 Outcome

What the Research Proved

It didn’t ship. There was no dedicated team to build or run it, and the agency was in the middle of a reorg at the same time, which made taking on something new infeasible no matter how much interest it had.

The team wanted this. Interest was never the obstacle, capacity was.

That’s not a footnote I’m hiding. It’s the reality of introducing a new discipline into an institution going through its own transition at the same time.

What it did prove is that research, not aesthetics, is what builds credibility for a new discipline inside an institution that doesn’t have one yet. Every recommendation I made had to be backed by a user’s own words or the agency’s own mission, because “this would look better” was never going to be enough. The archivist’s pain points carried real weight, because if the person who used this system every day was struggling, everyone with less familiarity was struggling more.

Ready when OHA isThe design specs, style guide and user flows are documented and ready to hand off when the agency has the capacity to build.

08 Reflection

Two Things I’d Sharpen Next Time

  1. Pressure-test the power users. I’d validate power-user workflows more deeply before consolidating search paths. I prioritized simplification without fully testing whether an expert user like Luci relied on multiple entry points for reasons I hadn’t uncovered yet.
  2. Speak leadership’s language. I’d build a harder business case. I focused on usability, not on quantifying what the current system was costing OHA in staff time or lost engagement. Design advocacy inside an institution moves faster when it speaks in numbers leadership already tracks.

Papakilo isn’t a generic database. It holds genealogical records and land documents that carry real weight for the communities who need them, and designing for that meant understanding why the content matters, not just what people need functionally.

01 Overview

A Database Unchanged Since 2009

2009the year Papakilo launched, and it had run essentially unchanged since
5people interviewed in depth: 3 internal staff and 2 external users
4 → 1separate search entry points, consolidated into one search bar
~5minto find a document in testing, down from 20–25 minutes

02 The Stakes

“What Is UX?”

Papakilo is the database Native Hawaiians use to access genealogical records, historical land documents, and cultural knowledge connecting them to their heritage. It had run essentially unchanged since its 2009 launch.

There was no UX function at OHA. When I proposed exploring a redesign as part of my UX certification, my team’s first question was, “What is UX?”

The real assignment

This wasn’t just a design problem. It was a chance to introduce user-centered thinking to a government agency that needed it and had no language for it yet.

An annotated UX audit of the Papakilo homepage and filtered search pages, covered in sticky-note callouts.
The first UX audit: key issue areas across the homepage and filtered search pages.

03 Research

If the Expert Struggles, Everyone Does

I interviewed 5 people. Archival research is a niche enough area that I chose depth over breadth.

Internal staff · 3

  • Luci MeyerCultural archivist, Papakilo’s daily expert user
  • Staff memberFellow multimedia designer
  • Staff memberBoard of Trustees staff

External users · 2

  • College studentResearching Hawaiian music history
  • First-timerNo archival research background at all

If the expert user struggles, casual users don’t stand a chance. Luci put it directly.

“There’s no way to help people keep track of information they want to keep referring to. They have to restart the search process every single time.”

Luci Meyer, Cultural Archivist

“Why are there four different ways to search? Which one gives me the best results?”

College student, external user

The external user hit the same wall from a different angle, unsure which of four separate places to even begin a search.

Illustrated portraits of the internal staff and external users I interviewed.
Who I talked to, and what each person needed from Papakilo.

04 The Real Problem

Not a Visual Problem. An Architecture Problem.

It wasn’t a visual design problem. It was an information architecture problem hiding inside a search problem. Fragmented entry points, no memory of a user’s search context, and filter categories that used internal jargon nobody outside the archive team understood.

I proposed consolidating four separate entry points into one search bar. I replaced the filtering system, which let users search only by collection, every single one listed out in a single dropdown, with four categories organized around how people actually look for things.

Search, restructured

4 places to start. Pick one before you know what you’re looking for.

  • Search by collection
  • Search by location
  • Search by newspaper
  • Advanced search

Filter by collection

…every collection, one long list, in archive-team terms

Low-fidelity user flow sketches exploring a single search entry point.
Hover the sketches to look closer. Early user-flow sketches for getting into search from one place.

05 The Proposal

Three Changes, Each Traced to a User

With the structure settled, I redesigned the screens around it. As the sole designer on this internal initiative, I owned the complete visual system and every design decision. I focused on three core improvements, and each one answers something a real person told me.

SOLUTION 01

One Search Bar Instead of Four Front Doors

Four competing search methods created decision paralysis. Nobody could tell which path would give them the best results.

The Old Design

Four different entry points (by collection, by location, by newspaper, advanced search) forced people to choose a search method before they knew what they were looking for.

The Proposed Design

One centralized search bar with progressive filtering. Start with a simple query, then refine as needed, so people can begin searching right away.

After: the proposed homepage with one central search bar.Before: the old Papakilo homepage with four separate search paths.OldProposed

Who this answersThe college student who asked, “Why are there four different ways to search?”

SOLUTION 02

Searches That Remember Where You Left Off

Even a power user like Luci had to restart her searches every time. There were no saved searches, no history, and no way to build on previous work.

The Old Design

Every search started from scratch. No memory of what someone had searched for, no saved parameters, no personalization.

The Proposed Design

User accounts with saved search parameters, search history, and recommendations based on the collections someone uses. Onboarding asks about research interests to surface relevant content up front.

After: proposed mobile screens with saved searches and history.Before: the old mobile search screens with no saved searches.OldProposed

Who this answersLuci Meyer, who watched people “restart the search process every single time.”

SOLUTION 03

Partners and Updates You Can Actually Scan

Dense blocks of partner information and poor visual separation made the interface overwhelming.

The Old Design

Database partners displayed as undifferentiated text blocks, with no clear way to scan for a specific organization or understand what each partner offered.

The Proposed Design

Partners shown as expandable cards organized by category, with clear hierarchy and breathing room between sections. “Latest updates” got more prominence so it’s easy to scan.

After: proposed partner cards organized by category.Before: the old partner page as dense blocks of text.OldProposed

Who this answersThe first-time user who said, “I don’t know where to start, so many things competing for attention.”

The Papakilo style guide: colors, type, buttons and components.
The style guide behind the proposal.

06 Testing + Refining

Four Testers, Five Minutes, No Help Needed

To check that the redesign worked and felt intuitive, I ran usability tests with 4 new participants, none of them from the original research group.

The task: find a specific historical document using the redesigned search and filtering.

Testing also surfaced three things to refine before I presented the work to stakeholders.

Before and after: the mobile menu with clearer labels and grouping.

REFINEMENT 01

Clearer Menu Pathways

Some testers weren’t sure how to reach certain features, so I improved the labeling and organization of the navigation structure.

07 Outcome

What the Research Proved

It didn’t ship. There was no dedicated team to build or run it, and the agency was in the middle of a reorg at the same time, which made taking on something new infeasible no matter how much interest it had.

The team wanted this. Interest was never the obstacle, capacity was.

That’s not a footnote I’m hiding. It’s the reality of introducing a new discipline into an institution going through its own transition at the same time.

What it did prove is that research, not aesthetics, is what builds credibility for a new discipline inside an institution that doesn’t have one yet. Every recommendation I made had to be backed by a user’s own words or the agency’s own mission, because “this would look better” was never going to be enough. The archivist’s pain points carried real weight, because if the person who used this system every day was struggling, everyone with less familiarity was struggling more.

Ready when OHA isThe design specs, style guide and user flows are documented and ready to hand off when the agency has the capacity to build.

08 Reflection

Two Things I’d Sharpen Next Time

  1. Pressure-test the power users. I’d validate power-user workflows more deeply before consolidating search paths. I prioritized simplification without fully testing whether an expert user like Luci relied on multiple entry points for reasons I hadn’t uncovered yet.
  2. Speak leadership’s language. I’d build a harder business case. I focused on usability, not on quantifying what the current system was costing OHA in staff time or lost engagement. Design advocacy inside an institution moves faster when it speaks in numbers leadership already tracks.

Papakilo isn’t a generic database. It holds genealogical records and land documents that carry real weight for the communities who need them, and designing for that meant understanding why the content matters, not just what people need functionally.

please feel free to browse other case studies below!

More Case Studies

Back to Projects
© 2026 Josh KohDesigned by me · Built with codeBack to top