2024
B2C • Web Application
Gemeente Amsterdam Veilig
Designing a safer, more humane way to connect citizens with government benefits
Duration
3 weeks
Team Size
4
My Role
Product Designer (UX strategy, research, interaction design, system logic, prototyping)
Client
Gemeente Amsterdam (Government of Amsterdam)
Agency Coach
Overview
This project was undertaken as part of my Master’s in Digital Design, in collaboration with Gemeente Amsterdam.
The brief was not to redesign a single interface, but to rethink how government benefits are distributed. Today, citizens must actively apply for resources, while eligibility is assessed through rule-based algorithms that often lack transparency, context, and empathy.
We asked a fundamental question:
How might government systems match resources with citizen needs in a way that is humane, transparent, and trustworthy, without increasing risk or complexity?
The Challenge
In the existing system:
-
Citizens must request benefits they are already entitled to
-
Algorithms flag potential fraud when criteria are not met, often without explanation
-
Decision-making lacks human context and empathy
-
Data usage is opaque, leading to low trust
-
Services are spread across multiple portals, increasing confusion and anxiety
Instead of optimizing a flawed process, we chose to rethink the system entirely.
As designers, we faced a second challenge:
Translating highly technical systems — algorithms, data permissions, and blockchain contracts- into interfaces everyday citizens could actually trust and use.
Discovery: What We Learned from Citizens
Through multiple citizen interviews, we uncovered consistent patterns:
-
People think in life situations, not benefit names
-
Managing multiple portals creates anxiety and errors
-
Fear of “doing something wrong” prevents applications
-
Lack of transparency feels punitive, even when outcomes are correct
This led to a critical decision:
Design one unified portal for all government support - centered on life context, not bureaucracy.
Solution Overview
We designed one citizen-first portal:
-
Life-situation entry points replace benefit names
-
Explicit consent and data-sharing transparency
-
Clear eligibility explanations and human escalation
-
Algorithm decisions are traceable via blockchain
Simple system flow
-
Citizen starts with a life situation (e.g. safety, housing, income support)
-
System requests only essential data
-
Eligibility rules are evaluated via smart contracts
-
Each decision step is logged on the blockchain
-
Citizen can see:
-
What data was used
-
Why a decision was made
-
Who accessed their data
6. Outcomes are either:
-
Approved
-
Requires clarification
-
Escalated to a human caseworker
This flow balances automation with human oversight.
Key Screens
Landing Page:
Sets expectations and builds trust before any data is shared.
Clearly explains how the portal works, how it differs from older government systems, how data is protected, and what other citizens say, reducing fear before first interaction.
Dashboard Day 0:
Explains system reliability, eligibility signals, and data use upfront.
Helps citizens understand what services are available to them and why, reducing uncertainty before starting any application.
Dashboard Day N:
Provides continuous clarity as applications progress.
Surfaces real-time status, decision reasoning, and system actions so users always know what’s happening, what’s next, and why - without needing to follow up or guess.
Key Decisions & Trade-offs
-
Rejected benefit-first navigation → chose life-situation framing
-
Rejected opaque automation → designed explainable outcomes
-
Used blockchain as trust infrastructure, not a visible feature
-
Prioritized clarity and ethical decision-making over speed
-
Explicitly designed out failure scenarios
Outcome & Validation
We stopped at concept and hi-fi prototype stage. To validate:
-
Tested with the same citizens from initial interviews
-
Presented at the MDD Feedback Feast, receiving input from visitors, faculty, and peers.
-
Measured whether the new flows solved real problems
.jpeg)
Observations:
-
Users easily understood where to start
-
Felt more confident navigating eligibility scenarios
-
Better understood why decisions were made
-
Reported increased trust in system and data usage
This early validation confirmed that life-situation entry points and transparent algorithmic explanations were effective design choices.
Failure Prevention
Design choices deliberately avoided common failure modes:
-
Prevented silent rejections → added explanation layers
-
Prevented algorithmic overreach → enforced human review
-
Prevented portal fragmentation → unified access
-
Prevented trust erosion → explicit consent & audit trails
Nothing “went wrong” because these risks were designed out early.
Algorithm & Ethics
Rather than hiding algorithmic decisions, we made them explainable, auditable, and traceable via blockchain. Humans retain authority in sensitive cases. The interface translates complex rules into clear, trustworthy guidance. Ethical decision-making and failure prevention were prioritized over automation for efficiency.
Key Learnings
-
Ethics must be designed in, not added later
-
Transparency builds trust faster than automation
-
Good UX reduces systemic risk, not just friction
-
Decision quality outweighs speed or pixel perfection
