My neighbor, Mrs. Robinson, is 70+ years old single lady who lives alone in her house. She belongs to the vulnerable demographic and her daily life has been greatly affected by quarantining. She fears to go out and often has anxiety while getting essentials from local supermarket. She is not tech-savvy and does not use delivery apps. Having worked & volunteered in the community, I know people who are ready to volunteer for delivering essentials to vulnerable people like Mrs. Robinson.

What it does

V-Delivery connects vulnerable demographics to volunteers, who help them by delivering requested essential items sourced from the local food banks.

  1. A requestor places a request for delivery of essential items on our toll-free helpline.
  2. V-Delivery receives requests and matches them to available volunteers. We notify the volunteer via text, and upon acceptance of the request we provide additional details for the delivery.
  3. A volunteer accepts a delivery request, picks up the items from the local food bank and delivers it to the requestor.

How I built it

The platform consists of the following components:

  • User Interaction - Requestors and volunteers interact with our toll-free helpline for intake of data. The helpline runs on a cloud contact center (Amazon Connect), that feeds the user interaction through a speech recognition bot (Amazon Lex) to understand and process the request.
  • Data Processor - After data intake, the data is processed by a cloud compute resource (AWS Lambda) that massages and stores the data in Google Sheets.
  • Match Generator - A workflow system (AWS Step Functions) processes available requests and available volunteers to generate a match. After a match is generated, the workflow notifies (AWS Pinpoint) the volunteer about the request details. Upon acceptance, a message callback workflow updates the match data and notifies the requestor about the acceptance.

Challenges I ran into

Our initial idea was to target senior demographics for delivery of any essentials they want. However, there are a couple of issues with that:

  1. Age verification - It is difficult, if not impossible, to verify the age over a phone call.
  2. Money exchange - Without face-to-face interaction, the exchange of money is difficult.

These challenges and short deadline made us pivot the idea by relaxing (removing) the age verification criteria and using food banks to avoid money exchange.

Accomplishments that I'm proud of

I learnt a lot about setting up a cloud contact center and integrating it with a speech recognition bot. Additionally, having floated this idea and PoC in my network drew a lot of interest.

What's next for V-Delivery

  • Closing the fulfillment loop - We want a way to know and verify request fulfillment.
  • Feedback & Kudos - We want a way to pass on the feedback and kudos from requestors to our volunteers.
  • Inventory Information - We want to provide food bank suggestions to our volunteers by integrating with their inventory.
  • Better Matching Algorithm
  • Partnering with local non-profits - After interacting with food banks, we learnt that there are other NPOs (like United Way of King County) that has a similar program.
Share this project: