← All work
Information Architecture · AI-Augmented Design

Living Bog Archive

Transforming fragmented cultural knowledge into an accessible digital archive.

Role
Designer & Researcher
Context
TU Dublin
Timeline
10 Weeks
Method
AI-Augmented Double Diamond
Living Bog Archive homepage, Our bogs. Our stories. Our future.
01
Overview

The Living Bog Archive is a cultural and educational platform documenting Ireland's raised bogs, landscapes that are ecologically vital and culturally rich, yet often fragmented across multiple sources and organisations.

The project explores how information architecture, editorial design and AI-assisted methods can help transform complex knowledge into an accessible and engaging digital experience, a community-led archive spanning heritage, biodiversity, oral histories, culture, research, events and conservation.

02
My Role

Designer and researcher throughout the entire project, from early discovery and structure through to a prototyped final design.

Research Information Architecture UX Design Content Structuring AI-Assisted Exploration Prototyping
03
Context

This project was developed as part of the AI for Designers postgraduate programme at TU Dublin.

The challenge was not only to design a website, but to explore how AI tools could be integrated into a human-centred design process, while keeping critical judgement and responsible decision-making firmly in human hands.

04
Problem Statement

Existing cultural archives often prioritise retrieval over understanding. Users can search for information, but struggle to see the relationships between places, biodiversity, stories, people and heritage.

The challenge was to create a structure that lets newcomers discover and learn, rather than simply search.

“The information exists somewhere. What's missing is a structure that helps people understand why it matters.”

05
Project Goals
Goal 01

Create a clear structure for exploring Irish bog heritage.

Goal 02

Support stories, biodiversity, culture and research in a single ecosystem.

Goal 03

Balance AI-assisted design with human judgement.

Goal 04

Design an archive that can grow over time without losing clarity.

06
Research & Discovery

I began by studying how cultural archives are structured and where they tend to fail their audiences, pairing that with desk research into the bogs themselves.

  • Analysis of existing heritage and cultural archives and their navigation patterns.
  • Information-architecture research into how people browse unfamiliar subjects.
  • Review of community-contribution systems that keep an archive growing.
  • Desk research into Irish raised bogs, ecology, history and cultural significance.
07
Key Insights
01

Users navigate through stories before metadata.

02

Discovery matters more than search for newcomers.

03

Cultural archives require context, not just content.

04

Community contributions are essential for long-term growth.

05

Information architecture is what turns information into understanding.

08
AI-Augmented Double Diamond

AI was woven through all four phases as an accelerator, never as the decision-maker. Each phase paired faster AI-assisted work with a human checkpoint.

Phase 01

Discover

AI: accelerated desk research, summarised heritage sources and surfaced archive patterns.

Human: verified sources and judged cultural relevance.

Phase 02

Define

AI: clustered findings and helped synthesise themes into a problem framing.

Human: set the problem statement and IA priorities.

Phase 03

Develop

AI: generated layout options, content structures and naming alternatives.

Human: chose hierarchy, tone and structure.

Phase 04

Deliver

AI: helped refine copy and stress-test the structure for gaps.

Human: finalised accessibility, editorial and design decisions.

09
Designing with AI

The clearest lesson was where AI adds leverage, and where a human must stay responsible for the outcome.

How AI helped

  • Research synthesis
  • Pattern identification
  • Idea generation
  • Layout exploration
  • Content organisation

What required human decisions

  • Accessibility
  • Cultural sensitivity
  • Information hierarchy
  • Community needs
  • Editorial judgement
  • Responsible AI considerations

Prompt engineering

Prompt quality directly affected output quality. Detailed, well-scoped prompts reduced unnecessary iterations, improved accuracy and avoided wasted AI usage. I structured prompts around a simple ICE framework:

Instruction
The specific task the AI should perform.
Context
The background, audience and constraints it needs.
Expectation
The format, tone and quality bar for the output.
sample-prompt · ICE
[Instruction] Suggest an information architecture for a public archive of Irish bog heritage.
[Context] Audience is newcomers, not researchers. Content spans places, biodiversity, oral histories, events and community contributions.
[Expectation] Return 3 structural options as labelled hierarchies, each with a one-line rationale and trade-off.

Sample prompt structure (placeholder), final IA decisions remained human.

10
Key Design Decisions
01

Why not a map-first homepage?

Although visually attractive, testing showed that newcomers lacked geographic context and became disoriented. A themed, narrative entry point led instead.

02

Why community contributions?

Local knowledge is central to preserving cultural heritage, and the only way an archive like this stays alive and grows over time.

03

Why a modular content system?

The archive needs to support stories, biodiversity, research, events and photography within the same ecosystem, without forcing each into the same template.

04

Why editorial storytelling?

Narrative structures improve learning and engagement, helping a newcomer understand not just what the bogs are, but why they matter.

11
Final Design

The final outcome: a scalable, reading-first archive spanning heritage, place, biodiversity, events and community contribution. Click any screen to view it full size.

12
Accessibility & Responsible AI

Accessibility-first

Inclusive design, clear navigation and strong readability built in from the start, not retrofitted.

Clear structure

A predictable hierarchy and consistent patterns so every audience can find their way without prior knowledge.

Human moderation

Community-contributed content is reviewed by people, keeping the archive accurate and culturally respectful.

Transparency

AI-assisted outputs are clearly understood as drafts, with final editorial decisions made and owned by a human.

Responsible AI

AI used as a collaborator within boundaries, accelerating work without outsourcing judgement or accountability.

Readability

Editorial typography and generous spacing make long-form heritage content calm and easy to read.

13
Reflection

AI is most effective as a collaborative design tool, not a decision-maker.

While AI accelerated research, synthesis and exploration, human judgement remained essential for accessibility, cultural sensitivity, community needs and meaningful information architecture. The result is a scalable digital archive designed to preserve and share Irish bog heritage for future generations.

Let's connect

Interested in how I approach problems like this?

Next project
STRIDE Checkout Flow