Goods intake that reads its own photos
Storekeepers already photographed every delivery into a chat. The agent turns those photos into an ERP arrival record and a set of publishable product images.
- Client
- Spare parts & industrial equipment · Germany · $15M revenue · 100 staff
- Scope
- Chat photos → condition → ERP → CMS
- Input
- Photos the team was already taking
- Role
- Solo — design, build, hand-over
Three people touching every delivery photo
The photos existed already — everything after them was manual cross-checking and sorting.
When spare parts arrived, the storekeeper checked them for damage, photographed them and dropped the images into a chat. A manager then cross-checked those photos against the delivery note and marked in the ERP that the goods had arrived.
Afterwards a second manager went through the same images by hand, picking the good-quality ones to publish on the website. The same photographs were reviewed three times by three different people.
One agent watching the chat
It reads the photos and the note, then updates both systems.
Photograph as usual
unchangedThe storekeeper photographs the arrived goods and sends them to the chat — the existing habit, unchanged.
Assess condition
visionThe agent pulls the photos from the chat and assesses the condition of the goods for damage.
Read the delivery note
ocrThe delivery number is read off the note and the delivery status is synced to the ERP.
Select and publish
cmsThe best-quality images are selected and pushed into the CMS, ready for the product page.
About five hundred hours a month
Measured against the three-pass manual process it replaced.
Saved every month
Passes over each photo set
Python