Method AI: Crafting an AI-Powered Recycling Assistant for Modern Workplaces
Designed and launched Method AI—an intelligent, location-aware recycling
assistant that reduced office waste contamination by 19 percent and laid the
groundwork for Method Recycling’s digital product ecosystem across New Zealand
and Australia.
What I built
* Multi-platform solution: iOS mobile app, web admin dashboard, and robust
backend.
* Real-time, AI-powered guidance for recycling at work—integrated with Method’s
physical bins and branding.
* Beta tested in ten workplaces, supporting facility managers and thousands of
office workers.
My role and tools
Role: lead product designer (ownership across strategy, research, user
experience, UI, and systems integration from launch to beta delivery)
Tools used:
* Figma: for design systems, wireframing, and hi-fi prototyping
* Xcode, Cursor: design-engineering handoff and app implementation
* Supabase: backend for AI RAG knowledge system and data storage
* CloudKit: iOS data synchronization
* Vercel: web deployment
The challenge
Despite clear visual guides, Method Recycling’s bins and signage could not
eliminate confusion—34 percent contamination rates exposed the need for smarter
support. Static instructions failed to adapt to region-specific recycling rules,
and busy employees made errors at the point of disposal.
Process overview
1. Problem and research
* Analyzed IoT bin data and identified peak confusion during lunch and snack
breaks; contamination baseline was 34 percent.
* Conducted user interviews and job shadowing to map pain points across three
core office personas: busy executives, sustainability champions, and
conscientious employees.
* Key insight: there is a three-second window to influence confident recycling
choices at the bin.
2. Design and prototyping
* Camera-first workflow: users receive point-and-dispose recommendations
quickly, minimizing friction.
* Progressive information: users get the “just right” answer fast, with the
option for detailed information and local council rules.
* Contextual guidance: AI adapts advice based on office location and specific
council regulations.
* Admin dashboard: facility managers and Method staff can track results, adjust
council guidelines, and review performance analytics.
3. Technical architecture
A layered, modular system ensures speed, adaptability, and scalability across
business environments.
Results and impact
* 19 percent reduction in contamination rates compared to pre-app baselines.
* 78 percent weekly retention among beta users.
* 92 percent correct identification on scanned items; 96 percent contextual
accuracy.
* 63 percent drop in AI processing costs using smart workplace caching, making
the platform scalable for business rollout.
* Established the technical and UI/UX foundation for further IoT bin
integration and corporate sustainability reporting.
Key learnings
* Ecosystem mindset wins: by tightly integrating the AI with bins and signage,
the experience felt native, accelerating adoption and confidence.
* Context matters: the RAG knowledge system built trust by delivering local,
council-specific advice instead of one-size-fits-all instructions.
* User behavior is shaped by moments: the three-second response target was not
just technical—it aligned with real workplace habits.
* Design for both people and organizations: the dashboard and admin tools
empowered both end users and facility managers, which is crucial for business
product success.
Takeaway
By understanding office behavior, leveraging AI for location-specific
confidence, and keeping everything fast, I transformed recycling from a source
of confusion into a moment of easy, positive action—setting up Method Recycling
for more impact and future digital growth.
