Dignitas has spent years building LeadNow, a platform their team uses to train and coach teachers across Kenyan schools. Through Tech To The Rescue they were matched with University College London's (UCL) Industry Exchange Network (IXN) for Good, and a team of five computer science students who spent two academic terms exploring how AI could support the coaching work. This was Dignitas's first tech partnership. UCL is a world-leading university and since 2011, the IXN has brought together students in Computer Science with external partners to devise technology proofs of concepts that solve real-life problems, as part of assessed student projects. In July, Sheena Visram and Graham Roberts from UCL joined Marita Adhiambo from Dignitas in our Impact Partnerships series to talk through how it went. The AI features are now in testing with Dignitas coaches, who are checking the feedback holds up pedagogically before it reaches teachers. Sheena has since reflected back on responsible innovation and the critical importance of working with charities as part of the IXN.
Artificial intelligence is often framed as a technological revolution, with discussions centred on models, algorithms and emerging capabilities. Yet in education, the most important questions are rarely technical. They concern people, pedagogy and how technology can strengthen, rather than replace, professional expertise.
This perspective shaped a recent collaboration between the University College London (UCL) Industry Exchange Network (IXN) for Good, Dignitas and Tech To The Rescue, which explored how AI could support teacher coaching in Kenya. The project was not just about building AI to demonstrate its capabilities. Instead, it examined how emerging technologies might enhance an already well-established system of instructional coaching, making expert support more accessible, consistent and scalable.
At UCL IXN for Good, we believe universities have an important role to play in addressing complex societal challenges. Beyond educating students, universities can serve as innovation partners, bringing together academic expertise, multidisciplinary student teams and organisations working at the forefront of social impact. This model creates space for responsible experimentation, allowing organisations to explore new ideas, generate evidence and de-risk innovation before committing significant resources to implementation.
"You understood both ecosystems. You understood where nonprofits have got strategic challenges, where technology could genuinely address a real-world problem and also the value of creating learning opportunities for students looking to understand the impact of emerging technologies in education." Prof Sheena Visram, speaking about Tech To The Rescue's role in the partnership
Our partnership with Dignitas illustrates the value of this approach. Dignitas had already developed a robust coaching methodology through LeadNow, its teacher development platform. Education Officers routinely collect rich classroom observation data, while teachers engage with professional learning resources and ongoing development activities. The challenge was not a lack of information, but the fragmentation of that information across multiple systems and processes.
The project therefore explored a broader systems question: could AI help connect lesson observations, coaching frameworks, teacher learning histories and curriculum guidance into a unified decision-support system? Rather than replacing the judgement of experienced educators, AI could potentially assist Education Officers by synthesising information, suggesting coaching prompts, identifying priorities for classroom visits and supporting evidence-informed follow-up conversations.
This distinction is significant. Much of the current discourse around AI focuses on automation, yet the greatest opportunities in education may lie in augmentation. Effective teaching depends on professional judgement, contextual understanding and strong human relationships. These qualities cannot be replicated by AI, but they can be supported through tools that reduce administrative burden, surface relevant evidence and make existing expertise more readily available.
The collaboration also reinforced the importance of designing technology around educational practice rather than technological capability. Early discussions deliberately shifted away from asking, "What can AI do?" towards considering the day-to-day decisions that Education Officers make before, during and after school visits. By grounding the work in authentic professional workflows, the technology became a means of supporting practice rather than dictating it.
For the students involved, this provided an equally valuable learning experience. Developing AI solutions in real-world educational contexts requires far more than technical proficiency. Students encountered practical considerations including connectivity constraints, multilingual environments, safeguarding responsibilities and varying levels of digital literacy. These realities demanded careful design choices and demonstrated that responsible AI development depends as much on empathy and contextual understanding as on engineering expertise.
The project also highlights the value of rapid prototyping within the social sector. Unlike commercial organisations, charities and education organisations often operate under significant financial and operational constraints. They must demonstrate value before making substantial investments in emerging technologies. Proof-of-concept projects offer an effective way to test assumptions, identify opportunities and generate evidence that can inform future funding, partnerships and implementation strategies.
Perhaps the most enduring lesson from this collaboration is that meaningful innovation is fundamentally collaborative. Universities contribute research capacity and fresh perspectives. Nonprofit organisations contribute deep contextual knowledge and practical expertise. Technology partners help translate ideas into workable solutions. Each perspective is necessary, and none is sufficient in isolation.
"Never think of it as handing over your product. Think of it as you are building together." Marita Adhiambo, Edtech Program Officer at Dignitas
As AI continues to evolve, its success in education will depend less on the sophistication of individual models and more on our ability to integrate these technologies thoughtfully into existing systems of professional practice. The goal should not be to replace educators with intelligent machines, but to equip educators with intelligent tools that enable them to do their work more effectively.
Ultimately, the promise of AI in education lies not in the technology itself, but in its capacity to strengthen human capability. When developed through genuine cross-sector collaboration and grounded in the realities of educational practice, AI has the potential to enhance teacher development, improve decision-making and contribute to better learning outcomes for students. That is a vision of innovation worth pursuing.
Sheena Visram is an Honorary Professor of Practice at UCL's Department of Computer Science and Co-Founder and Chair of the UCL Industry Exchange Network (IXN) for Good, which connects student teams with nonprofits and public institutions on societal challenges.
Sheena Visram and Graham Roberts from UCL joined Marita Adhiambo from Dignitas and our own Ciro Reynoso to talk through the collaboration, including what was harder than expected. The full session runs 45 minutes.
If you lead a social impact organisation and want to explore what AI could do for your work, our programmes are built for that. Find out more
If you work in tech or at a university and want your team on projects like this one, we can introduce you to organisations who need what you build. Get in touch
