ReClaim Vision

Grades a donated device for refurbishment — consistently — against a real refurbisher standard, and shows what it saw.

Closing the digital divide, faster. Refurbishing nonprofits (PCs for People, Human-I-T, Goodwill) receive donated computers by the pallet and grade them by hand — slow, inconsistent, volunteer-dependent. Inconsistent grading scraps good machines and ships bad ones.

Not an AI wrapper — a verifiable tool. ReClaim doesn't just ask a model "is this broken?" It grades against a versioned rubric grounded in a real refurbishment standard, shows the evidence it saw, and runs a deterministic safety rule in plain JavaScript: a device with a blocking defect can never be graded "refurbish," no matter what the model says. A volunteer can override any call.

1 · Capture the device

Upload a photo, use your webcam, or load a sample. Fill the frame; even, bright light; one device.

No image yet

Loading…

2 · Grade card

Capture or load a device, then click Grade device.

The rubric the model is graded against (and the exact prompt)