We are addressing the discrepancy between true demand and distribution-based demand of medical supplies to sites with poor internet access. The current model uses distribution-based demand, which is subject to inaccuracy due to lack of internet and other information systems infrastructure access.

Our solution attempts to solve this by:

  • using the Twilio API so that users can text orders instead of needing Internet access
  • creating a webapp with an interactive visualization of the demand of supplies by location

The webapp will inform Partners In Health (the supplier) as well as the hospital or clinic (the client) about demand trends and when to prepare for a higher discrepancy between true demand and distribution-based demand.

Made with a Flask backend, uses HPE Vertica as the database, visualization done with d3.js, and frontend uses JQuery and ajax. Deployed using AWS.

Try our app!

To visualize existing data points:

  • enter "tomsk" in location (a PIH site in Russia)
  • select "Neutralizer" or "Bleach disinfectant" in the webapp -- these are the most populated with data and return a fruitful visualization.
  • click Submit

To place an order and see it in the visualization:

  • text +1815-267-0042 with "4, KQ06, 20" [places order for 20 neutralizers -- sku KQ06 -- from location in Russia]
  • refresh the webapp
  • enter "tomsk" under location
  • select "Neutralizer" under "Item Category" -- you will see your new point to the far right of the visualization!

Future Steps

  • Implement functionality to allow placement of orders through webapp
  • Analyze timeline to estimate when the next order should be placed
  • Visualize item demand by location (by country, instead of just site) to determine interesting insights
Share this project: