STUDIO VIRIDIANDEVELOP · DESIGN · BUILD
VIRIDIAN RECLAIMS

Every material has a story — find its next one.

A just‑in‑time matching engine for the region’s salvaged, reclaimed, refurbished, surplus, and forthcoming‑deconstruction materials. Search what you need, see near‑fits you can modify, and compare each material’s benefit across the nine planetary boundaries of Kate Raworth’s Doughnut Model. Never a dead end.

Source Network

From the Viridian Partners Rolodex & CE Source Intelligence Log. Harvesters and forthcoming‑deconstruction tips feed the engine upstream; retail partners are the fast backstop. Call these before buying new.

How Material Match works — and where it's headed

What you see today is the interactive prototype, running on illustrative seed data. The cards below map each capability to how it becomes real — the "live data" build that turns this from a demo into Studio Viridian's sourcing backbone.

01 · AGGREGATE

From links to a live, unified catalogue

Right now the source network points outward. The live system pulls partner inventories into one searchable, deduplicated catalogue — with photos — and keeps it current:

  • Partner data pacts / feeds — Second Use, ReStores, Earthwise push stock on a schedule
  • Marketplace watchers — agents surface Craigslist / FB / OfferUp / eBay hits (their ToS forbids scraping, so a human confirms)
  • One “material passport lite” schema so every source normalizes to the same record + image
02 · SOURCE FROM YOUR SPEC

Google Sheet → AI sourcing agent

Sam's idea, exactly: a project bill of materials in a Google Sheet whose columns map 1:1 to these fields. Connect it and an AI agent runs a standing search + alert for every line, live‑updating the options it brings you to pick from.

  • See it staged in Post a project need → the spec‑sheet view
  • Automates the search & watch you'd otherwise post by hand
  • Design keeps deferred tolerances so late supply still fits
03 · CROWDSOURCE THE SUPPLY

Jobsites as the source — scan & incentivize

The frontier, per Sam: bring in JIT data straight off active jobsites. A staffer (then a trained AI agent) scans materials and auto‑tags them to our fields — turning a site into a marketplace with no middleman.

  • Prototype in the Jobsite capture tab; each confirmed scan trains the agent
  • Deconstruction: incentive is built‑in — sell on site vs. pay to haul
  • New construction: needs a regulatory / economic lever — tipping‑fee spread, diversion credits, EPR, decon ordinances
04 · DECIDE WELL

Cost ledger + all nine boundaries

Each match is compared against its virgin‑material alternative across a nine‑petal planetary‑boundary rose, not carbon alone, and, once sellers supply pricing, on delivered cost and avoided tipping fees. Environmental figures are illustrative estimates today; the live build replaces them with published data (EC3, environmental product declarations).

  • Per‑material and cumulative‑project views
  • “Optimize for” any single boundary
  • Built to balance trade‑offs, not hide them