Snap a photo. AI does the rest.
AI identifies each item from a photo, matches it to your catalogue and keeps a record of its condition.
For toy and tool libraries, food and baby banks, reuse shops and anyone who lends or hands out stock.
Intake
Photograph it instead of typing it
Use a phone camera or upload an image. AI suggests every field and ranks the closest items you already hold, with how sure it is.
- One photo and one tap, not 2–5 minutes of typing
- High-confidence matches confirmed in one tap
- New items arrive as a pre-filled form
Photograph an item
AI suggestions
Red Wooden Fire Engine
Handpainted wooden fire engine toy with ladder attachment, approximately 20cm long.
Colour
Red
Material
Wood
Condition
Good
Value
~£18
Minor paint wear on front bumper
Matches in your catalogue
Returns and condition
Know when damage happened, and who had it
Photograph a returned item. AI finds the loan, compares the photo with the original and flags new damage.
- Searches active loans only, so it finds the borrower
- A photo history for every item, from every intake and return
- One tap closes the loan and records the condition
Batch scanning
A box of donations in minutes
Take photos one after another with no wait between them. AI works through them in the background while you keep going.
- 10 items in under 5 minutes, against 20–50 by hand
- Matches and new items marked apart
- Confirm or adjust the lot in one go
Reporting
Know your impact
Each scan, loan and return feeds Plinth's wider reporting, ready for the impact reports funders ask for.
- Loans, active loans, items distributed and overdue rates, live
- What gets borrowed most, and what sits on the shelf
- Who borrows: postcodes, ages and repeat use
A cleaner catalogue
Duplicate detection
Fuzzy matching spots an item that already exists under a different name before you add it again.
Value from the photo
AI suggests a replacement value, so nobody has to look it up.
QR labels
Print labels for high-value or frequently loaned items to identify them instantly, without AI.
How the AI works
A vision model describes the photo by name, category, colour, material, size, condition and distinguishing features, usually in 1–2 seconds. That description is matched against your items and comes back as ranked matches with confidence scores. It works best at typical inventory sizes of 50–500 items.
Common questions
Who is AI stock tracking for?
Any organisation that lends, distributes or manages physical stock. Toy libraries track wear across many borrowers. Tool libraries log the condition of drills, saws and gardening equipment so they know when tools need servicing. Food banks batch scan whole deliveries and record what has been distributed to whom. Community libraries catalogue books, board games and DVDs, with QR labels for their most popular stock. Baby banks log the condition of pushchairs, car seats and clothing bundles, which are safety-critical. Equipment hire covers art supplies, sports gear and musical instruments. Reuse shops get value estimates for furniture, homeware and electronics and avoid duplicate entries. Schools and youth clubs see who has which shared resource and when it is due back. Community health teams track loans of mobility aids, sensory equipment and therapy resources, with condition assessed for safety compliance.
What does the AI fill in when we photograph an item?
Name, description, categories, tags, colour, material, condition with notes, and an estimated value. A high-confidence match to an item you already hold can be confirmed in one tap, and a new item arrives as a pre-filled form.
What happens to our photos and inventory data?
Photos and data are encrypted in transit and at rest, in line with GDPR. They are never used to train AI models. Every AI identification is logged, so you have a clear trail of what happened and when.