Federal ministry · Public Sector · Germany · 2024–2025

Redesigning Germany's Digital Funding Platform — for 20,000 users

Senior UI/UX design for the replacement of a legacy public-sector system — combining field research, service flows, prototyping and AI-supported requirements engineering.

Senior Designer Service Design UX Research AI Requirements Engineering WCAG Accessibility
Role
Senior UI/UX Designer
Client
National public administration
Duration
1.5 Years
Users
~20,000
~20K
Users on legacy system replaced
22
Design improvements delivered
AI
Requirements engineering — first on account
01Problem

Germany's public funding infrastructure was broken by design

Public funding in Germany was fragmented, opaque, and inaccessible — spread across disconnected portals and paper-based processes. The people most in need of funding support were the least equipped to navigate the system. Funding providers, meanwhile, lacked modern tools to publish, manage, or track programs at any meaningful scale.

The project required replacing a legacy system used by approximately 20,000 people with a platform that served two fundamentally different user groups with equally complex, often conflicting needs.

Funding Seekers

Citizens, SMEs, and organisations couldn't find, understand, or apply for relevant programs. Disconnected portals, no eligibility guidance, paper forms — the system excluded the people it was built for.

Funding Providers

Government bodies relied on manual coordination and paper across administrative levels. No unified tool to publish, manage, or track programs — creating inconsistency, delay, and compliance risk.

02My Role

What I was accountable for

I joined as a Senior UI/UX Designer, working across the web and app workstreams from research and journey definition through wireframes, prototypes and delivery. I collaborated closely with conception, engineering and public-sector stakeholders, and supported less-experienced colleagues with UX methods and design craft.

My contribution
Senior UI/UX design across web and app workstreams, from discovery through detailed interaction design
Introduced client workshadowing at regional operations office — a new research method for this account and client
Supported junior conceptionists with scrum practices, wireframing and UI/UX foundations for day-to-day delivery
Co-introduced AI-native requirements engineering — structured epics and user stories maintained by AI agents in VS Code, replacing manual Word/Confluence documentation
Facilitated workshops and customer interviews, documented decisions and kept design deliverables moving during a temporary team absence
03Process

Research → Flows → Prototyping → Testing → Iteration

Phase 01Observe & Understand

Immersive field research — including workshadowing

Conducted a structured kick-off with intensive touchpoint mapping. Ran qualitative interviews with both Funding Seekers and Providers. Introduced workshadowing at regional operations office — spending time alongside customs officials to observe real workflows and capture requirements that interviews alone would have missed. This was a first for the engagement.

user-journey-map-v2.fig
FUNDING SEEKER SYSTEM Awareness Discovery Eligibility Application Tracking Decision Can't find it Paper forms Landing page Search & filter Eligibility check ✦ Digital form ✦ Status tracker ✦ ✦ New in redesign Pain point identified in research
Phase 02Ideate

Synthesis into design directions

Translated 20+ research findings into clear UX requirements. Synthesised qualitative insights from interviews and workshadowing into design directions that addressed both user groups simultaneously — a structural challenge unique to dual-sided platforms in regulated environments.

Phase 03Envision

Co-creation with real users and providers

Ran validation sessions with Funding Seekers and government providers to test concept direction and prioritise features. Balanced user-centred insights against strict federal compliance requirements — designing to WCAG accessibility standards from the first prototype.

lo-fi-wireframes-v4.fig
Search & Discovery Suche nach Förderprogrammen… KMU Bundesland Branche Eligibility Diagnostic ✦ Sind Sie ein KMU? Ja Nein Frage 2 von 5 Voraussichtlich förderfähig Provider Dashboard 247 Anträge 18 Offen 94% Bewilligt Antragsteller Status Müller GmbH Bewilligt Bauer AG In Prüfung
Phase 04Refine & Deliver

22 design improvements across 8 research cycles

Iterative delivery across sprint cycles — each improvement grounded in structured testing. Built and maintained design documentation in close coordination with development. Introduced AI-native requirements engineering to maintain epics, user stories, and feature roadmaps at scale — the first time this approach was used on the account.

04AI Use Case

Why AI — and not a classic approach

Manual requirements documentation in Word and Confluence had been creating a persistent bottleneck — slow to update, easy to desync from actual development state, and dependent on a small number of people to maintain.

// AI-native requirements engineering — what changed
Requirements are now stored in a structured, version-controlled format in Visual Studio Code and maintained by AI agents. The AI creates and updates epics, user stories, and feature roadmaps based on agreed standards — removing the manual documentation burden and keeping requirements in sync with development at all times. The goal: AI-generated UI flows and screens in the future, closing the loop between requirement and prototype automatically.

This was the first time this approach was introduced within the consultancy’s public-sector practice — co-established with the development team and rolled out progressively across the project.

05Key Decisions

Where the hard design choices were made

Decision: One platform or two?

Chose a single platform with role-based views rather than two separate products. Harder to build — but critical for shared data, consistent compliance logic, and administrative coherence across Funding Seekers and Providers.

Decision: Eligibility check before or during application?

Moved eligibility screening to the discovery phase — before users invest time in a full application. Reduces invalid submissions, builds trust, and surfaces the right programs earlier. Validated in user testing across multiple sessions.

Decision: Accessibility as foundation vs. retrofit?

Committed to WCAG 2.1 AA from the first wireframe — not as a QA step at the end. This required early alignment with development but prevented the costly rework that accessibility retrofits always demand.

Decision: AI requirements tooling — introduce now or later?

Introduced AI-native requirements engineering mid-project rather than waiting for a clean start. The transition required investment but returned immediate speed and consistency gains — and created a model the team could carry into future engagements.

06Solution

What we built and delivered

Smart Search & Discovery

Intelligent filtering matching seekers to relevant programs based on sector, size, region, and project type — replacing the keyword search on the legacy system.

Eligibility Diagnostics

Guided pre-application flow that surfaces eligibility before users invest time in formal submissions — reducing invalid applications for providers and frustration for seekers.

Dual-Role Interfaces

Separate, role-tailored interfaces for Seekers and Providers — each designed around distinct goals, permissions, and task flows — unified under one platform.

Accessible by Default

WCAG 2.1 AA compliance built from the first prototype — ensuring the platform serves all citizens equally, including those with disabilities.

Provider Programme Management

Tools enabling government bodies to publish, edit, and track funding programs at scale — without technical dependencies or manual coordination across teams.

Application Status Tracking

Real-time visibility for Seekers into application status — reducing the uncertainty and inbound follow-up burden on both sides of the platform.

07Collaboration

How the team worked together

Client (Federal authority)
Conducted workshadowing sessions with operational staff and presented research findings and design directions to stakeholders across administrative levels.
Engineering
Tight day-to-day alignment with development on feasibility, component specifications, and AI requirements engineering. Design informed sprint planning directly.
Conception (L9–11)
Supported junior conceptionists with UX methods, wireframing and scrum, helping embed more consistent design practices in daily delivery.
Delivery leadership
Worked independently within the senior design team and helped maintain continuity of workshops and deliverables during a temporary team absence.
08Impact

What changed — and what it enabled

~20K
Users on legacy system being replaced
8
Research sessions & interviews conducted
22
Design improvements delivered across iterations
1st
AI requirements engineering on account — first of its kind
09Learnings

What worked, what we adapted

Workshadowing unlocked requirements that interviews couldn't. Spending time alongside customs officials at regional operations office surfaced workflow constraints and edge cases that no interview guide would have reached. Proximity to real work is irreplaceable.
Designing for public sector means holding two things simultaneously. User-centred thinking and strict legal/federal compliance are not opposites — but they require active negotiation at every design decision. The best work finds the overlap.
Delivering insight is political in regulated environments. Knowing how to frame a research finding for a civil servant vs. a developer vs. an delivery director is its own skill — and failing to adapt the framing means insights don't land, regardless of their quality.
AI requirements tooling compounds over time. The upfront investment in establishing AI-native requirements engineering paid back in consistency and speed within two sprints — and created a reusable model for future public-sector engagements.