Back to Case Studies
Data-driven Web Application to Address Racial Inequity in Schools in the USA

Case Study

Data-driven Web Application to Address Racial Inequity in Schools in the USA

Challenge

Development of a web platform that automatically manages and presents mass survey data aiming to address the racial inequity children face in schools in the USA.

Skills

Web Development

Year

2020

About

Discriminology is an American Start-up aiming to leverage technology, data, and human capital to advance the social, civic, and economic strength of Black and Brown communities by ensuring their youth have equal opportunities for success through equitable opportunities to learn.

The Discriminology Plus platform aims to address the racial inequity children face in schools in the USA. The product enables schools and parents from the community to easily undertake surveys with students and instantly see the results in an intuitive dashboard.

The data visualisation approach of the project allows users to see details about the survey scores and a list of recommendations for tackling any possible racial inequity issues.

Marzee team was responsible for developing the Minimum Viable Product (MVP) to digitize the survey process that the Discriminology and Village of Wisdom (VoW) teams were doing manually.

We worked closely with Major, who was responsible for all the Product Design and UI/UX processes.

Cover image by unsplash

Tech Stack

The platform was developed based on the Serverless Application Framework model, built to auto-scale using AWS product offerings.

The key components used are DynamoDB, Cognito, Lambda and API Gateway. The surveys are powered by TypeForm and fed to the platform via webhooks. Everything is served in the front-end with ReactJS.

Project Highlights

Marzee intervention in the project started after Major delivered the Design phase. We already had access to all the user stories, user flow and design. With all of this in hand, developing the tech architecture that would best fit the product was easy.

From here, our team could organise the whole project agile delivery in 2 weeks sprints, enabling client feedback throughout the project duration. It allowed us to adapt to the product’s needs and change the development course accordingly.

  1. Survey

    The product's main goal was to allow users to manage surveys with students. From the beginning, we knew that for an MVP, we should not develop the whole survey process from scratch. A better strategy was to use well-established third-party services. It made development smoother and kept maintenance effort and costs at a minimum.

    We had two main options to use for the survey process: SurveyMonkey and TypeForm. Both of them are well-supported products with great Documentation, so we could easily work on top of them.

    Considering the nature of the product, we initially thought that using SurveyMonkey would be a better fit, especially because it allowed for further customization. However, when further comparing pricing and overall offer, we quickly understood that TypeForm had a much more reasonable pricing plan.

    The modular way we’ve built the platform, where all the pieces are decoupled and glued together by AWS, allowed us to make this shift instantly.

  2. Serverless Infrastructure

    We followed a serverless approach to build the Discriminology Plus product. This means that we used a cloud service, in this case AWS, to manage server provisioning, scaling, and maintenance.

    This also allowed us to use a series of different tech and third-party services and serve everything in a React front-end with Amplify (the AWS Content Delivery Network).

    More than that, by relying on AWS services, the product was built to scale easily while controlling costs.

    The following AWS services are being used here:

    • Cognito for all the user authentication and management
    • DynamoDB to store and manage survey data
    • Lambda functions to trigger all the different pieces together
    • API Gateway to create and manage all the product's API by connecting everything in one application.

    This infrastructure was key to keeping the client in control of the running costs and allowed us to create the application in a modular approach. We developed smaller modules and components combined to build the complete Discriminology Plus application. This allowed for faster development cycles, quicker feature deployment, and shorter time-to-market for new features and improvements.

    More than that, this approach made it easier for us to adapt to new needs for the project, considering the 2020 pandemic. For instance, the product was initially thought to be used only by School Leaders, but the Discriminology team realized it was important to include other stakeholders as users. We were then able to include Community Leaders as part of the audience for the product by creating a new type of user role.

  3. Dashboard

    Beyond running surveys, another important aspect of the Discriminology Plus product is the presentation of the survey results.

    School Leaders and Community Managers with an account can start a survey, send it to students and close it as soon as they want. Once they do it, they can instantly see the survey results in an intuitive dashboard with key metrics and scores.

    React Front-end development for Data Visualisation

    For each metric, users can see a breakdown of the answers by race, gender and grade. Allowing the school and community managers to have a detailed understanding of how the students are feeling about those topics.

    More than that, it is possible to see the demographics of the survey to have a better understanding of the data sets.

    Depending on the survey results, the School Leader or Community Member has access to a series of recommendations prepared by Discriminology team.

    Serverless infrastructure with AWS services

    We used Recharts to build the charts in the dashboard. It’s a good React library all-around: it has good documentation (both for APIs as it’s built with TypeScript and code usage). It handles different types of representations (Pie, Circular, Bar, Scatter etc.), which makes it a good choice for the future. Furthermore, it’s built on solid grounds (as it’s using both React and D3 under the hood).

Results

The overall development of this MVP was a very interesting process. We had a lot of fun working with the Discriminology and Major team.

  • The client was very happy with the result and started using the product immediately within their community.
  • Our development modular approach allowed for flexibility and adaptation during development to meet their business needs.
  • Having all the designs, user-flows and stories ready by the time we started development contributed to a focused and agile development.
  • The Serverless infrastructure we built was key to controlling costs and delivering a fast and more flexible platform.
  • High-quality development, communication and security standards throughout all steps of the process.