AWS Public Sector Blog

Happy Sad app leverages AWS generative AI to improve student well-being

AWS branded background design with text overlay that says "Happy Sad app leverages AWS generative AI to improve student well-being"

The COVID-19 pandemic took a heavy toll on students’ mental health and well-being. In fact, a staggering 87 percent of public schools reported that the pandemic had negatively impacted their students’ social-emotional development during the 2021–2022 school year. These effects have lingered long past the pandemic, and students’ social-emotional well-being remains a primary concern of administrators, teachers, and parents. To address this ongoing crisis, The Happy Sad Company was founded.

“We designed the Happy Sad app based on our direct experiences as parents, teachers, and longtime tech entrepreneurs,” said Hannah Oldknow, company founder and a former data scientist for Instagram. “The goal of our platform is to help answer the question: ‘How is my child doing?’” Oldknow envisioned an app that would help teachers, parents, and students gain a better understanding of students’ social-emotional well-being.

Collaborating with Amazon Web Services (AWS) for the strategic planning, scaling, and launching of the app was the clear choice, according to Oldknow. The Happy Sad Company significantly accelerated app development by taking advantage of AWS advanced tools, including Amazon GuardDuty, AWS Lambda, and Amazon CodeGuru. And the team was impressed with the work AWS has done with other education technology (EdTech) companies, such as Metis and Gaggle.

Using artificial intelligence (AI) to engage students and empower teachers

The Happy Sad app serves a dual purpose. When K6 students access the app, they enter a gamified experience accented with cloud-based emojis and personalized accessories. The goal is to help students improve their emotional self-awareness, a process supported through the application of social-emotional strategies based on the well-known CASEL framework.

The app’s other purpose is to provide educators access to individual students’ “mental health report cards,” which are based on their activity and responses in the app. The teacher can view this data to identify patterns and trends in the emotional status of individual students—or across the entire classroom. AI insights derived from the data can also be exported and shared with parents.

Eventually, the founders envision the app will serve as an early notification system for students’ social-emotional well-being, alerting teachers and parents if a student’s emotional health begins to decline.

“We’re now training our proprietary AI so it can predict drops in a student’s baseline, which would help parents and teachers get ahead of a potential problem,” said Oldknow.

AWS expertise accelerates app development

Robust support from AWS at every step helped The Happy Sad Company move quickly through app development. The company was able to use the broad experience AWS has with other EdTech companies to help guide its prototype development process.

“Part of the reason we teamed up with AWS was knowing that they could guide us on where the challenges would be, especially around compliance,” said Oldknow.

From strategic planning to scaling and launch, a cross-functional team at AWS provided deep expertise and support to Happy Sad throughout their app development process.

“We held a lot of workshops with the Happy Sad team, helping them better understand which AWS services made the most sense for them to use as they built their app,” said Daniel Wells, an AWS solutions architect.

Along with solutions architects, a variety of other AWS teams shared their expertise along the way, including the startup Activate team, account management team, go-to-market team, and the Worldwide Public Sector team.

AWS guidance proved especially helpful in navigating the array of AI-powered tools available to improve the app development process. With two patent-pending native AI programs to its credit, The Happy Sad Company’s experienced AI team quickly saw the value in adopting AWS technologies. The team decided to use CodeGuru Reviewer, a tool that uses AI to suggest ways to optimize the app’s code, as well as GuardDuty to oversee security and monitor for malicious in-app behavior.

Prioritizing compliance while reducing costs

Care for the well-being of students is at the heart of education. With the growth of educational technology, that care now extends to students’ data privacy. The Children’s Online Privacy Protection Act (COPPA) places requirements on websites and online services (including apps) to protect the privacy of children under 13—a priority The Happy Sad Company shares.

“The Happy Sad team was focused on making their app secure and compliant,” said Leo Zhadanovsky, chief technologist for education at AWS. “Our teams provided detailed guidance on meeting compliance requirements wherever it was needed.”

By prioritizing COPPA-compliant app development, AWS only strengthened its close partnership with The Happy Sad Company.

“Leo’s team did a fantastic job of setting up guardrails to maintain data privacy, which we need to maintain the trust of our users,” said Oldknow.

The Happy Sad app required a flexible, scalable infrastructure that could seamlessly adapt to demand spikes without extensive management overhead. AWS serverless computing services proved to be the ideal solution. By using the serverless computing platform AWS Lambda, the team bypassed costly capital expenditure investments in fixed hardware. Instead, they were able to instantly tap into the virtually unlimited pool of secure compute resources that AWS offers, paying only for the capacity they actually used. This serverless approach, as Zhadanovsky explains, ensures “you’re never over-provisioned. Auto scaling means accessing exactly the power you need, when you need it.”

Bringing the Happy Sad app to market

The Happy Sad app is rapidly gaining adoption as a powerful classroom resource for K6 student mental health. App usage has quickly spread among teachers across the United States, with larger districts, including Los Angeles Unified School District, Houston Independent School District, and Columbus (Ohio) City School. Resource engagement among teachers varies between 50–70 percent, while engagement among student users now averages about 80 percent month to month.

“Parents are excited to see the direction the app is going in,” said Oldknow. “Many have been wanting a tool that can provide greater awareness of their children’s mental health. We’re excited to deliver it.”

Read related stories on the AWS Public Sector Blog: