Data and AI

Global Data Science Challenge 2025: Green agents of change

As part of our global partnership with UNICEF, the Global Data Science Challenge 2025 contributes to UNICEF’s Green Rising initiative, a bold effort to mobilize millions of young people across the world with the skills and opportunities to build a sustainable future.

This year’s challenge invited thousands of changemakers at Capgemini to harness agentic AI and turn climate anxiety into climate action.

Amazing things happen when people and technology collaborate to make things better, and often all that’s needed is a way to determine where and how to begin. For the 2025 edition of our annual hackathon, supported by Capgemini’s technology partners Amazon Web Services (AWS) and Mistral AI, we challenged participants to develop an agentic AI-powered assistant that can help young people explore green learning and career pathways.

This initiative reflects our shared commitment to advancing the United Nations Sustainable Development Goals, particularly climate action and quality education. By combining UNICEF’s mission to empower youth with our expertise in AI and data, the challenge demonstrates how technology and purpose can work together to create lasting change.

Agentic AI made real

Climate change is a growing concern for young people worldwide. According to joint research by the Capgemini Research Institute and UNICEF’s Generation Unlimited, Youth perspectives on climate: Preparing for a sustainable future, 67% of youth fear for the planet’s future, yet only 44 % feel equipped to act.

With the aim of tackling this situation, the Global Data Science Challenge (GDSC) was designed to close this gap by transforming concern into capability. Participants were asked to create an agentic AI assistant that could guide young people toward green learning and career pathways. To support this, we provided curated datasets including green job descriptions, climate training resources, and youth personas. Teams had to combine technical innovation with human-centered design, ensuring their solutions could:

  • Understand individual aspirations and goals
  • Identify relevant skills and learning opportunities
  • Recommend personalized career paths in sustainability

This approach not only tested advanced AI techniques but also emphasized ethical design and inclusivity, ensuring that technology serves as a tool for empowerment rather than complexity.

GDSC 2025 teaser

Green agents of change

Winning solution: The Green Career Assistant

Over 1,500 participants formed 300 plus teams and submitted more than 3,000 solutions. Among these, the winning team from Germany, oCaptainMyCaptainPlanet, stood out with its neuro-symbolic AI assistant that:

  • Cleans, organizes, and interprets raw data to uncover insights, and build structures based on international frameworks
  • Engages with users to understand their needs, pain points, and goals so it directly addresses each user’s actual needs
  • Uses a knowledge graph to link data, ideas, and competencies to create smarter connections
  • Ensures humans remain in control, including the ability to adjust final recommendations.

In addition, the winning team designed the solution to track its own energy consumption and carbon emissions and then optimize its operation to make it more sustainable.

This solution was selected by a jury of experts from UNICEF, Capgemini, and our technology partners along with a youth representative. The solution is now being considered for potential pilot deployment in Brazil, where it could help test real-world impact.

Enabling innovation through AI

This year’s challenge was made possible through the support of Capgemini’s technology partners. Mistral AI provided advanced AI models and solutions that enabled participants to build more adaptive and efficient AI assistants. AWS powered the challenge with scalable cloud infrastructure and AI services, helping teams develop and deploy their solutions with speed and reliability.

Together, they formed the technical backbone that allowed participants to focus on building solutions that matter. This collaboration accelerated innovation during the hackathon and laid the groundwork for future advancements in sustainable AI. GDSC continues to show how partnerships can drive meaningful impact at scale.

Driving impact beyond the hackathon

The Global Data Science Challenge is more than a competition – it is a platform for learning, collaboration, and innovation that continues to inspire real-world change. By connecting youth empowerment with advanced AI, GDSC 2025 demonstrates how technology can accelerate progress toward a sustainable future. The insights and solutions developed this year will inform future initiatives and partnerships, ensuring that the momentum for climate action and green skills continues to grow.

UNICEF does not endorse any company, brand, product, or service.

About the Capgemini Global Data Science Challenge

Launched in 2016, the Capgemini Global Data Science Challenge has evolved into an annual, purpose-driven hackathon that brings together thousands of Capgemini team members from around the world. Participants delve into a real-world data science or AI use case, honing their skills while collaborating with seasoned professionals and non-governmental organizations. GDSC has become a proven driver of impactful innovation, enabling thousands to learn how to apply advanced technologies to make real progress towards addressing some of the world’s most pressing environmental and social challenges.

Client stories

Enabling evidence-based education decision-making with agentic AI

Capgemini teams up with UNESCO and AWS for the 7th edition of the Global Data Science Challenge to develop an agentic Gen AI system that can empower countries to make informed, evidence-based policy decisions to enhance foundational learning outcomes

Read more

Harnessing the power of AI to combat river blindness

Capgemini partners with the University Hospital Bonn to bring an end to River Blindness. As part of this effort, they launch the fifth Global Data Science Challenge focused on developing an AI solution to help with diagnosis and the development of new treatments for the disease

Read more

Deepening our knowledge of the ocean with AI and machine learning

The LoVe Ocean Observatory can now quickly sift through the massive amounts of data it gathers from oceanic sensors to better inform the conversation around marine conservation

Read more

Using AI to protect whales

Working with Lisa Steiner, Capgemini and AWS develop a new, self-learning AI solution that can more effectively identify matches between

Read more

Learn more

Catching the AI bug

Identifying insects through AI-driven audio analysis. The winners of the Global Data Science Challenge 2023 share their experiences of the competition, which involved developing an AI-driven model to identify insect sounds.

Code for a cure

Using AI to treat river blindness – a solution that demonstrates the ability of AI to effectively diagnose the stage of worm development within the patient, as well as machine learning’s capacity to speed up the review process needed for new treatments.

Love the Ocean

Fighting climate change with undersea intelligence with the aim of developing a concept AI solution to help the Love team review and analyze their data more efficiently.