WRMX-01

shAIring – Turning one ordinary apartment block into a tiny circular economy.

Comment

Inspiration

By 2030, cities will be denser, stuff will be cheaper, and climate pressure much higher.
But everyday life in a normal apartment block will still look the same:

  • everyone owns a drill, a party set, piles of clothes – used once, then forgotten
  • most of it ends up as waste instead of being shared or reused

We asked: what if one ordinary building could behave like a tiny circular economy – by default, not just for eco-nerds?

What it does

shAIring is an AI helper for a single apartment block.

Residents don’t fill forms, they just talk to it:

  • “I need a drill tomorrow.”
  • “I want to get rid of old clothes and books.”

shAIring then:

  • matches people who need things with neighbors who have them
  • creates and tracks borrowing agreements between residents
  • proposes swap / collection events when several people want to declutter
  • estimates how much CO₂ and waste the building saves by sharing instead of buying new

On top of that, it keeps everything transparent with:

  • a building dashboard (residents, items, events, impact)
  • a borrowings overview (what I borrowed / what I lent)

Challenges we ran into

  • Making it feel like a future city service, not just a chatbot
    The main endpoint is /api/chat, but we had to design the system so it changes the shared reality of the building (items, events, impact), not just send pretty messages.

Accomplishments that we're proud of

  • Built a working AI-driven backend that:

    • interprets free-text requests
    • decides an intent (borrow, dispose, ask impact, etc.)
    • and updates the shared in-memory building state accordingly

  • Implemented a clean React + shadcn UI with:

    • chat interface
    • building dashboard
    • borrowings overview

  • Designed a simple but meaningful impact model:

    • every borrow and event translates into approximate CO₂ and waste savings
    • residents can see their building’s impact grow over time

  • Created a concept that clearly maps to the hackathon megatrends:

    • AI Everywhere – embedded in the building, not just in people’s phones
    • Climate & Green Transition – reuse by default, not as an afterthought
    • Urbanized Life & Trust – anonymous neighbors become a sharing community

What we learned

  • How to orchestrate an LLM as a tool-using agent, not just a text generator:

    • strict JSON outputs
    • clear intents and actions
    • a backend that remains the source of truth

  • That small UX decisions matter:

    • “I need a drill tomorrow” feels very different from a form with 5 fields
    • showing impact in simple numbers makes climate benefits feel real

Built With

Share this project:

Updates