Our presentation!

Inspiration

  • Buying Items online has terrible implications for the environment:
  • A 2500 Mile Shipment(SF-NYC) * 2lbs Item weight * .0006595kg CO2 Emitted per lb-Mile = 7.2697431 lb CO2
  • Emitted to ship that item
  • Over 3x your item’s weight emitted in CO2!
  • Online shopping isn’t going away, and most people don’t pay attention to where their items are shipping from eBay provides the standard convenience/assurance of online shopping but also lets you choose to shop locally

What it does

  • ecoBay helps redirect you from shopping sites like Amazon, where items are sometimes shipped from halfway around the world overnight to you, to eBay results local sellers.
  • This has a massive reduction on the carbon footprint of your order, with little inconvenience to you! (The eBay listings often even save you money!)
  • We also analyzed our commit history via MLP and plotted some graphs to tell our keystone development story!

How I built it

  • We used NLP powered by Google Cloud to parse the site for keywords that describe the product and sent them to our Django backend on GCP, which called the eBay API and returned closer items. A user can then scroll through the eBay options and pick the best product.
  • We also leveraged esri to help visualize the impact/CO2 savings gained by shopping local/via ebay. We use an esri map and geolookup to compare/map you location and that of your eBay purchase, vs. that of the closest Amazon Warehouse or location your item would have been shipped from.
  • We also use UiPath to automate bug/outage notifications to the developer both to their computer directly, and to their email.

Challenges I ran into

  • Having to use pure JS (no frameworks) was a pain factor for those of us with a lot of JS experience in node,react,native,etc.
  • Ran into a lot of trouble getting UiPath to work, got it done though!
  • Albert’s computer keeps bluescreening!
  • Had to work over discord instead of in person due to even cancellation, loss of face to face time is always difficult.

Accomplishments that I'm proud of

  • The NLP product recognition is working really well. It both has a decent success rate, and seems to work on most sites/product pages tested
  • Our ebay queries are working perfectly
  • App is fully functional/has a utility in current state
  • UI is clean/simple
  • Successfully got esri and UIPath working
  • Our first ever browser extension!!
  • Proud of our development story/the NLP

What I learned

  • Just how massive the carbon footprint of Online shopping/shipping items are
  • How to make a web-browser add-on (Google Chrome Extension in particular)
  • A lot more about working in pure JS
  • The ins and outs of the ESRI platform
  • The power of GCP and its NLP API
  • eBay’s API and the amount of local products available
  • UIPath

What's next for ecoBay

  • Tweak NLP even further for better results
  • Map more of the Amazon warehouses
  • Improve integrations/dataflow between out APIs
  • Port to Firefox
  • Live carbon pricing/calculation
Share this project:

Updates