Inspiration
Recycling all our lifes, love for nature and a good business understanding
What it does
Categorize Scrap to help evaluate and sort it.
How we built it
Writing Python for most parts aand Javascript for the frontend. We deployed everything in AWS. But most and for all with love.
Challenges we ran into
Lots of unusable images, incomplete data, complicated data without further documentation. Too many categories.
Accomplishments that we're proud of
A pretty good image categorization model. A good inventory forecast.
What we learned
That even with 10'000 images it is worth it to have a glimpse at it and figure out quite something. How to extrapolate from incomplete data.
What's next for What the Scrap
Improving picture quality by working closely with Thommen Inc. Getting more data from SAP with documentation and maybe a developer. Look for material prices on the net.
Log in or sign up for Devpost to join the conversation.