Project status · Work in progress. Navigator is an ongoing, exemplary side project developed together with a backend engineer as a shared learning exercise.
Role
Designer, Prototyper with AI - testing out what AI can and can't do well, Collaborator working with two backend engineers
Context
To explore AI as technology in the design process as well as the topic of AI I choose a topic that spans audio and compliance. Both topics have been par of my previous employment at Apple Supporting Logic Pro and at Valispace helping aerospace engineers manage traceability of requirements. AI as technology impacts the work and industry in Audio / Music as well as in Systems Engineering. Audio andn Systems Engineering workflows adjust to manage AI specific and non AI specific compliance. Inspired by these shifts, I made a simple prototype addressing compliance in the sound and music industry. Since AI can be used to produce or enhance sound, companies have been swamped and Laws are drafted while everything figures out what to do with AI. A source of truth that helps access and digest laws and regulation reliably could help simplify
Result
After working on this case for my portfolio, I saw a call for Mozilla Fellowship submissions and submitted the project. I passed the first round of evaluations and invited two backend developers to work on the project. Currently the navigator runs as a side project and is being refined
Navigator Interface
Use Cases
- Consent & Attribution — Was this work used with permission?
- Training Data Provenance — Where did training data come from?
- Generated Output Rights — Who owns AI-generated work?
- Transparency & Disclosure — When must I disclose AI use?
- Authenticity & Recognition — How can creators prove and protect original work?
Use cases lean creator-side — individual artists typically want guidance for their specific scenario, not a full compliance audit.
Documentation
Handover & Collaboration
Handover & Collaboration
Prototype handed over to a backend engineer for professional implementation. Ongoing collaboration on: tool logic and decision-tree structure, auto-updating regulations from official sources, accessibility and language translation (my responsibility). This is a mutual learning exercise — we each teach the other what our discipline knows.
AI + human: where each does what
This project is deliberately honest about where AI helped and where I made the calls. The split below is the actual working arrangement, not a retrofit.
AI-assisted
- Requirement gathering across scattered regulatory sources
- Cross-verification across models (Claude, Grok) to catch single-source errors
- Rapid React prototyping for the decision-tree UI
- Scaffolding drafts for the developer-facing handbook
- Reformulating legal text into plain language candidates
Human-led
- Design direction, information architecture, editorial decisions
- Audience framing — who reads this, at what reading age, in what context
- Source verification against primary regulation texts
- Accessibility review (WCAG 2.2 AA), tagged structure, alt text
- Reconciling tensions — e.g. creator-protective vs. developer-facing framings
- Engineering collaboration, handover decisions, scope control