AWS for Industries

Introducing Breaking Barriers Initiative: Building generative AI applications for digital inclusion

Breaking Barriers Hackathon title card

With the advent of generative AI services such as Amazon Bedrock and Amazon Q Developer, it has never been easier to generate high-fidelity code while also building applications with foundation models. With this ease of use comes both great opportunity and responsibility to use this technology to impact critical barriers that exist today across digital inclusion, social equity, and sustainability.

AWS for Telecom is a dedicated global business unit working with operators across the globe. Through our community we can connect hundreds of millions of people, and we’re proud to have the opportunity to do so. That is why we are pleased to announce Breaking Barriers: an innovative worldwide program designed to accelerate the development and deployment of generative AI solutions aimed at promoting digital equity. Through cross industry co-innovation, including regional hackathons, the initiative aims to crowdsource ideas and design artifacts for frictionless, secure, and sustainable frameworks that integrate leading generative AI models with connectivity solutions, creating universally accessible digital experiences.

In this post, learn about our first-of-a-kind hackathon designed to accelerate the development of generative AI solutions aimed specifically at promoting digital equity.

Our first Breaking Barriers Hackathon

Hosted in partnership with industry leaders such as Amazon Web Services (AWS), AT&T, NVIDIA, and TM Forum, the four-day event challenged teams to use generative AI in creating accessible, equitable digital solutions across verticals such as education, healthcare, and customer service. With $17,500 in cash prizes, the hackathon doubled down on transformative ideas that could translate into commercially viable products and services. The hackathon harnessed the latest advancements in AI: from Amazon Bedrock running Claude 3.5 Sonnet and Amazon SageMaker, to partner solutions such as NVIDIA NIM and ODA Canvas. Moreover, it aimed to empower underserved communities and unlock new avenues for business growth.

With just five weeks of planning, we amassed almost 600 participants and 80 project submissions from teams across three continents for the in-person event in Plano, Texas. Given the unforeseen capacity, we instituted a first round of submissions to select the top 34 projects to compete in-person in front of our prestigious judging panel:

Kaniz Mahdi: Director of Technology, Telecom Business Unit,– AWS
Mark Austin: VP, Data Science – AT&T
Julian Williams: Telecommunications Industry Lead – Anthropic
George Glass: CTO – TM Forum
Fred Devoir: Global Head of Solution Architecture and Engineering – Telco, NVIDIA

In just 72 hours, the teams built solutions across numerous industries, including:

Healthcare – digital nurse providing real-time diagnoses and explanations of esoteric medical domain knowledge
Education – virtual teacher that could distil complex subjects into a multitude of languages to meet the needs of diverse learning abilities
Retail – application for real-time clothing pattern customization and immersive product interaction
Travel & Hospitality – Virtual reality (VR) application to allow aging population to interact with a virtual avatar as they “travel” to new locations

I’m pleased to announce the winners, who claim portions of the $17,500 prize:

First Prize: Nexxus – Composed entirely of UT Dallas students, the team created a personalized education platform (LearnAI) that rendered real-time games to facilitate comprehension that met the user’s specific learning abilities.

Figure 1 LearnAI application user interface, developed by Tousif Ahsan, Manvir Chakal and Mohammed Siddiqui

Figure 1: LearnAI application user interface, developed by Tousif Ahsan, Manvir Chakal and  Mohammed Siddiqui

Second Place: Arcee Labs – This team built a solution to help democratize artificial intelligence/machine learning (AI/ML) technologies to help underrepresented communities by using AI to create AI. They created a solution that used an Agentic AI framework to build domain-specific models for languages such as Arabic, Hebrew, and Swahili, and learning differences such as ADHD and Hyperlexia. It enabled organizations to easily create models that are specifically tailored to reach a wider audience and foster inclusivity by breaking down existing language and learning barriers. Their solution also used an agentic base model trained to accept natural language input to make decisions and orchestrate the build, training, and deployment of tailored models for each of these communities.

Figure 2 AI creating AI architecture

Figure 2: “AI creating AI” architecture developed by Chris Smith, Lucas Atkins, Jerome Lintz and Julien Simon from Arcee Labs

Third Prize; 63NationalParks – As another UT Dallas student team, racing back from a final exam to the AT&T Innovation Center to present their solution, they developed an interactive Massive Open Online Course (MOOC) application that uses generative AI to index lecture videos and provides high-impact query interventions sourced from students of diverse backgrounds, dynamically tailoring responses based on the user’s knowledge level. This created a vibrant social classroom experience that fostered deeper understanding and engagement.

Figure 3 Application user interface designed by Rohith Peddi, Likhitha Pallapothula, Bharath Challa and Pratik Manghwani

Figure 3 – Application user interface designed by Rohith Peddi, Likhitha Pallapothula, Bharath Challa and Pratik Manghwani

Over the coming weeks, we’ll be collecting project submissions and developing an open-source repository of solutions to democratize access to many of the use cases built above. Stay tuned!

Conclusion

It’s an incredibly exciting time as we help customers define what generative AI can mean to their business operations, their customer experiences, or even their service portfolio as they look to expand to adjacent verticals.

The solutions showcased at our first Breaking Barriers Hackathon demonstrate the incredible potential of generative AI to drive tangible impact, whether that’s improving accessibility, boosting productivity, or unlocking new sources of growth. To learn more about Breaking Barriers, visit AWS for Telecom or the Breaking Barriers homepage to find the next event in a city near you.

To stay up-to-date on the latest events and announcements, register on our interest form or reach out if your organization would like to partner as a co-host, Lightning Talk Speaker, or even judge!

On behalf of AWS for Telecom, we challenge all of you to continue pushing the boundaries of what’s possible. Keep innovating, keep collaborating, and let’s make the promise of digital equity a reality, together.

Robert Belson

Robert Belson

Robert is a Developer Advocate in the AWS Worldwide Telecom Business Unit, specializing in AWS Edge Computing. He focuses on working with the developer community and large enterprise customers to solve their business challenges using automation, hybrid networking and the edge cloud.