Back

Case study

Harnessing data and ethical AI to improve efficiency in sustainability education

Reboot Education is the flagship initiative of UK non-profit Reboot the Future. It supports 22,000 teachers with free resources aligned to the national curriculum, designed to foster the values and aims of the UN Sustainable Development Goals (SDGs). 

As part of a tech-for-good collaboration powered by Tech To The Rescue’s accelerator programme, Opencast worked with Reboot the Future in summer 2025 to harness data and ethical AI to help take the Reboot Education platform to the next level.  

An Opencast team of data, AI and machine learning specialists designed and delivered a proof of concept to demonstrate how Reboot can streamline content management and improve the discoverability of resources across its platform.  

This work forms the foundations of a powerful platform that will make it easy for educators to find the content they need – and allow Reboot to keep better control of its own data. 

2hrs>5 mins 

Time taken for labelling

80-90%

Time saved using AI

80%

Accuracy on labelling  
(figures based on proof of concept)

Challenge

Meet growing demand with limited resources

As Reboot’s library of content has grown along with increased demand, so has the challenge of managing and evolving it. With a small team and a deep commitment to sustainability and ethical technology, Reboot needed to add technical expertise to help it move on to its next stage. 

First, Reboot needed to solve some critical technical problems: how to manage more than 1,600 content resources more efficiently, improve content discovery for its users and ensure any solution remained ethical, low impact – and aligned with the charity’s values. 

As demand for Reboot’s content had increased, the charity’s platform struggled to keep pace. Built on a Django backend, the site hosted hundreds of educational resources for teachers. But  managing these materials required manual effort from a small team.  

 

Each new resource had to be categorised by key stage, subject area and relevant Sustainable Development Goals (SDG) themes – often without consistent metadata or formatting.  

Beyond content management, users faced issues with search and discoverability. Teachers often couldn’t find what they needed quickly or intuitively, which limited the platform’s ability to support classrooms in real time. Additionally, the environmental and financial costs of scaling infrastructure were top of mind. The organisation needed a solution that was lean, smart and future- ready. 

To ensure that AI was viable for the organisation, the team needed to make sure that the underlying data was both good quality and well-structured – but there were challenges in how they operated their educational platform, including: 

  • Manual content management  
  • Limited discoverability  
  • Sustainability concerns. 

Reboot was a participant of the AI For Changemakers programme, an initiative from Tech To The Rescue to help organisations use AI to multiply their impact. In the programme, Reboot had access to a guided AI bootcamp, mentors and important materials – but needed the right technical partner to test whether AI could genuinely improve their operations before committing to a larger transformation.  

Response

Harness data and AI to amplify human insight

With a strong track record in public services and a deep commitment to tech for good, Opencast saw Reboot’s challenge not just as a technical problem – but as an opportunity to support meaningful social impact. The collaboration was built on a foundation of shared values: transparency, empathy, sustainability and a belief that technology should empower people, not overwhelm them.  

Together, Opencast and Reboot set out to explore how AI could be used not as a replacement for human insight, but as a partner to amplify it. 

The primary objective of the project was to design a first proof of concept (POC) that could automate content classification, improve resource discoverability and reduce the operational overhead for Reboot’s team. The team progressed to build the content classification element of the job, with the wider POC serving as a blueprint for future development of the Reboot AI in Education platform. 

Three clear goals guided the project: 

1. Automate resource classification using custom AI models to categorise content by: 

  • Key stage (early years through key stage 5) 
  • Subject area (25+ labels) 
  • SDG-aligned thematic topics (16+ labels). 

2. Build a scalable, serverless processing pipeline using AWS tools to extract, classify and update resources – starting with PDFs, which made up the bulk of Reboot’s content. This included full data governance and lineage metadata. 

3. Establish an ethical, low-impact architectural foundation that could grow with the platform while aligning with Reboot’s sustainability principles. 

Opencast’s data specialists worked closely with the Reboot team from day one, holding discovery workshops to understand the team’s workflows, priorities and ethical constraints.  

The team built a scalable, serverless AI-powered system that automates the classification of thousands of educational PDFs, significantly reducing manual workload and operational overhead.  

This foundation, supported by robust deployment automation and clear documentation, enables seamless integration with Reboot’s Django platform and allows for rapid growth and iteration. 

The approach sought to ensure that ethical AI could meet scalable infrastructure. The way we work emphasises transparency and knowledge sharing, ensuring the Reboot team remained in control of their platform and informed about each decision, – helping ensure long-term self-sufficiency. 

The complete proof of concept was delivered to its agreed scope in just four weeks – two weeks ahead of schedule. This extra time allowed for thorough testing, user feedback and several rounds of refinement. 

Technical details

  • AWS Textract

    was used to extract text from PDFs, enabling automated processing of Reboot’s content. 

  • AWS Lambda

    functions orchestrated the end-to-end flow and data governance, allowing for event-driven, scalable processing without managing servers. 

  • AWS Comprehend

    models were trained with Reboot’s own labelled data to perform high-accuracy, multi-label classification across age groups, subjects, and topics. 

  • AWS CloudFormation

    the architecture was implemented entirely as Infrastructure as Code using AWS CloudFormation. 

  • GitHub Actions

    for automated deployment and CI/CD. 

The result was a fully serverless AI classification pipeline, underpinned by a data platform on top of object storage using Amazon S3. This is capable of ingesting and processing content at scale while keeping infrastructure minimal and cost-effective. 

Impact

Easier access to meaningful, purpose-driven content

The work overall has helped Reboot progress towards an intelligent, scalable data-driven platform powered by AI. Key benefits of the solution are: 

  • high-precision classification, improving the relevance of search results and resource suggestions 
  • automated updates to resource metadata, reducing the manual workload for Reboot’s team 
  • a future-ready architecture with full governance, easily extendable to new content formats and more advanced features like adaptive learning paths 
  • low environmental impact and operational cost, aligning with Reboot’s sustainability commitments. 

With Opencast’s guidance, Reboot is now aware of technologies and solutions that it can use at no cost, and this enables the charity to keep better control of its own data. 

Perhaps most importantly, the process itself empowered Reboot to understand and steer the role of AI in its platform. Opencast provided guidance on AI ethics, cost control and data governance, ensuring Reboot could grow with confidence. 

The project has paved the way for enhanced user experience, along with better retention and alignment with sustainability goals. Above all, it has been a collaborative effort grounded in shared values and a clear goal – to make it easier for educators to access meaningful, purpose-driven content.  

The success of the project has opened a new chapter for Reboot the Future. With a strong technical base and a deeper understanding of what’s possible with AI, underpinned by a strong data foundation, the organisation is actively developing its next-generation Reboot AI in Education platform. It will offer: 

  • personalised learning experiences for teachers, guided by AI-powered recommendations based on user needs and classroom dynamics 
  • collaborative teacher networks, where educators can exchange resources and support each other through challenges like eco-anxiety and burnout 
  • tools for student empowerment, including campaign frameworks and lesson plans aligned with the SDGs 
  • resources for wellbeing, using storytelling and small, hopeful actions to help teachers and students build resilience 
  • a scalable loyalty model, where organisational clients remain engaged through purpose-driven learning and development. 

By August 2025, Reboot was preparing to integrate the AI pipeline into its live platform, explore new personalisation features and expand access to more educators across the UK and worldwide.  

Reboot was also engaging with potential funders and partners to support the continued development of the platform and its in-person components. 

The partnership between Reboot the Future and Opencast is a testament to what’s possible when vision meets capability – and when technology is used not to replace human insight, but to amplify it. 

The partnership has shown that scalable data and AI solutions can be built ethically, affordably and sustainably. More importantly, it has laid the foundation for a platform that doesn’t just distribute content, but sparks change. 

As Reboot continues its mission to help educators and organisations build a better world, Opencast will remain a committed partner. As part of the wider tech for good community, both organisations aim to lead by example – to prove that, with the right people and principles, technology can truly help reboot the future. 

The project has reinforced the value of collaboration, trust and shared purpose in delivering meaningful technology. It has also highlighted the growing potential of harnessing data and ethical AI in education – especially when shaped directly by the people who use it. 

For Tech to the Rescue, this collaborative project is one of many powerful examples of what can happen when non-profits and technologists join forces to serve the common good. 

  • AI-powered data-led content classification

  • Enhanced search functionality for content discoverability

  • Lean, sustainable cloud-based solution

  • Validated and scalable AI integration

As our library of resources and demand from teachers grew, we needed a way to manage content more effectively without compromising our values of sustainability and ethics. Opencast’s expertise in data and AI helped us design a solution that not only streamlines content management but also makes resources more discoverable for educators. The approach was collaborative and values-led, giving us the confidence to harness AI as a force for good.

Anthony Bennett, Chief Executive

Reboot the Future

Download case study

Fill out the short form and we’ll email you a link to download a PDF of the case study.

*
*
*

Related Content

Loading...