Numiko’s HE Hackday

The Challenge: Finding the right course is a pivotal moment for prospective students, but the traditional course search tools on university websites often assume that students already know exactly what they want to study. For many, that simply isn’t the case. Team 1 set out to create a more engaging and inspirational course discovery process, one that would support students who might not yet know their perfect path.

The Solution: The team's answer was a survey-driven course finder that lets students discover potential courses based on their interests, qualifications, and aspirations. Rather than relying on filters or predefined categories, the tool engages users through a series of dynamic questions that lead to tailored course recommendations. What's more, the prototype incorporated additional resources, such as alumni success stories and potential career paths, helping students to visualise their future more clearly.

Key Insights: Team 1 found that successful course discovery requires a deep understanding of student psychology as much as it does technical innovation. Students want to be inspired—they’re looking for more than just a list of course options.

The team also recognised the importance of considering an international audience, where qualifications and educational backgrounds vary greatly, making personalisation even more critical. This exploration highlighted the challenge of making course discovery truly global and inclusive.

We thought this tool could be particularly for prospective students who aren’t completely sure if they want to go to university and are exploring their options.

The Challenge: Universities have large and complex digital estates. This is a challenge we’re intimately familiar with, working with organisations like University of Oxford, who manage over 800 different websites. Content managers face challenges uploading and maintaining all this content, but digital tools can help.

Team 2 took on the challenge of making this process easier and more efficient for higher education institutions, exploring how AI could streamline some of the ‘boring stuff’ that makes up day-to-day content management, freeing up CMS admins to create more and better content.

The Solution: Rather than focusing on a single tool, the team looked at how AI could automate lots of little annoying CMS tasks that slow content admins down. These included tools for auto-tagging and content categorisation, text tone analysis, SEO optimisation, and even translation services for international students. By building prototypes, the team demonstrated how these AI tools could reduce the burden on content creators while ensuring that the quality and relevance of the content remained high.

The text tone analysis tool, shown below, helps content admins maintain a consistent tone of voice by reviewing their copy against their brand's tone of voice characteristics. It can rewrite the copy to match the desired tone or let admins edit it themselves.

The SEO analysis tool below uses AI to analyse the page's content and provide recommendations for making it more SEO-friendly. This could help HE teams increase the likelihood of their course appearing in SERPs for a relevant search term.

Key Insights: The team found that while the technology exists to automate much of the content management process, integrating these tools into real-world workflows is a complex task. Many of the AI modules they tested were still in development, and there was a lack of cohesion between the different functionalities.

Nonetheless, the potential for AI to asset content managers is clear, and the team identified several areas for future improvement, including better user research to understand where the greatest pain points lie.

The beauty of the approach taken by team 2 is that it lends itself to rapid implementation. The fact that these tools merely offer suggestions to admins, who can rectify any AI generated mistakes prior to them being seen by users means the risks of using AI in these contexts are minimal.

The Challenge: In an age where Gen Z are accustomed to quick, attention-grabbing digital content, universities face the challenge of making their often lengthy and detailed pages, such as course pages, both engaging and relevant. Team 3’s brief was to create personalised experiences that would resonate with prospective students used to short-form, dynamic content.

The Solution: Team 3 decided to tackle this challenge by developing a custom GPT model that could personalise the content shown to students based on their previous browsing behaviour on the university website. Rather than relying on intrusive pop-ups or lengthy surveys, the system would track what students were looking at on the site and then generate a personalised page tailored to their interests. This page would display relevant content without overwhelming the user, allowing them to engage with only the information that mattered to them.

The team also explored the potential of AI video, which could use the content generated by the custom GPT to offer personalised videos tailored to individual students.

The prototypes below are for illustrative purposes only and are not intended to represent the University of London.

However, the team decided that AI video couldn’t offer the short-form and quick-paced content needed to fit the brief. Instead, they focused on creating a personalised page for users to give them the key info they'd need, given the profile developed by their previous browsing behaviour. This page would feature content personalised to the user selected by the AI model.

Key Insights: One of the team's biggest takeaways was the importance of refining AI prompts to ensure they produce the right output for the target audience. Personalisation, especially when aimed at Gen Z, needs to be done in a way that feels natural and helpful, rather than forced.

This project highlighted how AI can help users without being obtrusive when a bit of creativity is applied to its implementation.

The Challenge: Choosing a university isn’t just about the courses on offer— it’s about the whole experience. Many students consider factors like the local environment, social life, and opportunities outside of the classroom when making their decision. However, university websites tend to focus only on the academic side of things, neglecting the wider context. Team 4 set out to solve this problem by finding a way to showcase the unique advantages of a university’s location.

The Solution: Initially, the team explored using 3D virtual tours to capture the campus environment, but they soon realised this wasn’t enough. Instead, they pivoted to creating an interactive recommendation tool that would suggest local activities, places to eat, and other experiences based on the student's preferences. By pulling in personalised recommendations, the tool would give prospective students a real sense of what life could be like both on and off campus, helping universities stand out in a crowded market.

The prototype below is for illustrative purposes only, and is not intended to represent Ravensbourne University.

Key Insights: The team realised the need to pivot when they realised their initial line of enquiry (3D virtual tours) didn't solve the key problem for universities. That kind of quick and creative exploration of new tools is the entire point of hack days, and there's nothing wrong with taking a few wrong turns when exploring a new tech landscape.

Their recommendation tool makes university websites far better at showcasing the value of their place, a factor that is key to students selecting a university, but is often given little thought by university websites. To produce a working version of the tool, the team would need to gather more data from current students to ensure the recommendations are both authentic and relevant. This stage of building the tool would be a great way to engage existing students and leverage student-generated content.

Our Higher Education Hack Day uncovered some exciting opportunities for creatively applying tech to solving real problems.

By bringing a wide range of varied skill sets together, we were able to explore different viewpoints and gain a better understanding of what's possible. One of our PMs said "The working tech that the devs can spin up is mind-blowing!" and this captures the whole point of the day. By learning more about what's possible we can shape and propose new, more effective solutions to the challenges faced by our clients.

Going forward we'll be refining these ideas, gathering feedback, and continuing to push the boundaries of what's possible in this space. If you're a university looking at utilising technologies like those showcased here to boost student recruitment, make life easier for your content admins, or inspire more young people, get in touch. We'd love to hear from you.