On Track Learning creates tailored and engaging training courses which are contextualised around relevant stories and scenarios. From business compliance to health and safety, and people management courses, their learning platform provides every type of training teams need to thrive and grow. Their courses are created to be industry specific, using in-context storytelling to make the learning more memorable, help employees build their knowledge and apply it to their jobs.
With the On Track Learning team, ASquared ideated, designed and developed the new Novella system. Users input a fact sheet (or ‘course bible’) and specific parameters (job title and business context) to generate SCORM compliant courses that deliver learning structured to Bloom’s Taxonomy (a set of hierarchical models used for the classification of educational learning modules). All of the courses created are story-based to aid learning which is then tested with multi-format questions. Regardless of skill level, anyone can create multiple eLearning courses within minutes with built-in error checking in place, plus zero deviation or hallucination possible. Novella AI revolutionises personalised and contextual learning, at scale.
With over 500 Learning Management Systems (LMS’s) available, On Track Learning needed to stand out and showcase their unique position in the market. Whilst they had an incredibly personalised approach to course creation, the process was time consuming to deliver at scale. After initially beginning as a website brief, our team strategised alongside the On Track Learning leadership team.
We ideated a radical approach to integrating AI technologies in a clever, slick and innovative way to modernise their position in the learning landscape and quickly scale their reach. The learning and development industry can be quite slow moving and is absolutely ripe for digital disruption. ASquared love working with clients like On Track Learning who are already owning their market but where we can add our digital expertise into the mix to help supercharge and grow a business exponentially!
Having previously relied on third party platforms, there was a huge opportunity to design and develop their own platform to enable the creation of bespoke, contextual eLearning content utilising generative AI technologies. Across an evolving project, there later became a need to create HTML learning course modules which were SCORM compliant, and to build a production-ready version of the initial web prototype complete with AI images, video, and more. This would allow On Track Learning to take their new product to market at scale and pursue business investment opportunities.
ASquared initially set out to create an AI powered prototype to auto create text and image content modules with multiple choice question integration. As the project progressed, two stand alone projects were defined: an AI content authoring tool, and a bespoke SaaS Learning Management System.
We followed a user-centric design and production process, with the outcome being a new bespoke web application - Novella. Our aim with version one was to build a product, which could be tested with a segment of new and existing users, innovated upon and further developed into a solution that ultimately would replace the current offering. The new system was built to produce learning modules that were not only compliant with existing Learning Management Systems but also auto generated by AI in minutes rather than taking multiple days to manually author.
This phase kicked off with stakeholder workshops, and product and process reviews. We achieved an in-depth understanding of our client’s business goals, landscape, customers and the product requirements. New product thinking accompanied a deep technical exploration and tech spikes of various AI technologies before we established a technical approach and began feature planning. First, we created lo-fi UX and UI journeys. We brought the prototype to life using a strong design direction and clear visual language through multiple one-week long iterations which included an initial user interface design and associated component library, created by a cross-functional design and development team.
We worked to embed generative AI technologies into the core of the new course authoring tool, in this case OpenAI ‘completion’ APIs, chosen for high quality output, solid SDK, excellent documentation and scalable service. With this generation stack, we distilled On Track Learning’s learning ethos, and automated the creation of their unique contextual learning courses. The authored course content can then be automatically transcoded via Synthesia AI into multiple AI video modules, which the user then has the option to check and edit before publishing. This creates an automated, personalised, accessible and accurate course with incredibly engaging delivery that caters for all learner types.
The initial goal of phase two was to implement an export to standalone, HTML learning course modules with compliance (a set of technical standards for eLearning products). We investigated SCORM, xAPI and TinCan, and concluded with progressing SCORM compliance. We conducted a short tech spike to align on the best approaches to future-proof the business, whilst ensuring backwards compatibility with existing and legacy systems - SCORM is almost 30 years old. Through our tech spike, we identified opportunities to craft modern and responsive, yet still SCORM compliant, learning courses that are usable on any device. During this phase, the authoring part of the system remained a prototype, but the newly generated exported files were production ready.
With the goal of enhancing On Track Learning’s current process for creating courses using AI tooling, we created a production ready AI authoring tool. Taking the existing proof of concept, we created a service that runs the prompt engineering tooling on remote servers, allowing fact sheet inputs and self-checking processes to run independently of the website. We further refined the AI model prompting, to achieve both creative output and factual correctness. In parallel to the development work, we iterated on the design of the user interface. The course generator was also upgraded, making editing and publishing the course content more fluid and intuitive for the user.
We then enhanced the course generation model to leverage AI image and video generation with programmatic Synthesia video avatar integration and contextual Dalle-3 images. This added further engaging elements to the existing text-based delivery. The AI generates usable image libraries on the first try, with the tool allowing for additional images and media to be inserted into the course CMS. In the final development element, we iterated the fact sheet mechanic to include a library of previous course inputs, and allowed for new fact sheets to be uploaded from a simple template to create courses on any topic. Lastly, a new option for course generation was added for the AI not to tell a story, just to order the facts and test them.
The new web-based platform - Novella - builds bespoke, contextualised eLearning courses in minutes, rather than days. Our AI-enabled learning platform empowers HR managers to dynamically create industry-centric, scenario-based learning programmes, addressing HR challenges with efficiency and effectiveness and making irrelevant, tick-box eLearning a thing of the past. The production solution was highly scalable, utilising backend service tools like Clerk for user authentication and Supabase for data management.
The result is a groundbreaking, disruptive digital product which utilises AI in eLearning in a completely unique and industry-leading way.
.png)
Through our network, we are also helping On Track Learning to pitch Novella to angel investors and VC’s to accelerate their growth, marketing abilities and future development. We’re looking forward to continuing to work with the amazing team and pushing the boundaries of what’s possible in this space.
"We approached ASquared and in our very first meeting they told us our tech was fine, we don’t really have a problem or immediate need…but, we do have some innovative opportunities and ideas we could collaborate on.
They then built a prototype which totally blew our minds and we now have a working commercial product. When I shared it with the Digital Design Specialist at a large international enterprise they said, 'this is the most advanced use of AI I have seen so far in learning technology.'"
