Skip to content

The tech behind LXP de-mystified

The LXP is a platform on its own, but functions best when acting as a portal for a number of enabling technologies. Here are the ones you need to understand, simply explained.

You don’t have to be a hands-on coding genius to get value out of a learning experience platform for your organization. However, it helps to have a working understanding of the key technologies that enable it to do what it does. 

Most people who own and operate a learning management system (LMS) will know at least a little about SCORM, the open technology standard that enables an LMS to package and track elearning modules. But the emergence of the learning experience platform (LXP) has brought a variety of new enabling technologies which might be less familiar to the average learning professional.

If you belong to one of the 50% of companies who research tells us are likely to buy an LXP in the next 24 months, it really helps to know something about these underpinning technologies. Otherwise, how will you know what you are buying? What the systems you are looking at should do, can do – and just as importantly, can’t do.

Researching these technologies can be a time-consuming, bewildering and even frustrating business. LXP is an emerging market space, which inevitably means that information available on the web is likely to be partial, fragmentary and in some cases contradictory. The terms and acronyms used to describe things can vary from one source to another, while some of the information will be just plain wrong.

In order to speed your learning about the technologies that underpin LXP, we’ve assembled this small but perfectly formed glossary of terms, written as much as possible in plain English.

What do these technologies enable an LXP to do?  

For most of us, the important thing about any technology is not so much what it is, as what you can do with it, in practical terms. To give a context for this glossary, it’s useful to start with a definition of what a typical LXP does, in order to explain how the underpinning technologies enable LXPs to perform those functions. 

‘LXPs are single-point-of-access, consumer-grade systems composed of integrated technologies enabling learning. They can do many tasks, such as curating and aggregating content, creating learning and career pathways, enabling networking, enhancing skill development, and tracking learning activities delivered via multiple channels and content partners.’

Janet Clarey, Bersin by Deloitte

Learning Experience Platform Glossary of terms

CMS (content management system)

A content management system (CMS) is software that supports the organization, modification, and presentation of digital information. If you’ve ever used WordPress to publish a personal blog, or a web builder like Squarespace, you are using a content management system.

At the heart of almost all learning systems is content management technology. A learning management system (LMS) is, arguably, just a very specialized type of CMS. However, the classic LMS was designed to handle a very limited range of content types; principally, SCORM-packaged elearning modules and documents. As bandwidth and computing power increased over the years, graphics, animations, video, games and many other content types moved into mainstream use for learning, and the LMS has struggled to keep up. Many people built learning portals (content managed websites) to supplement or front-end their LMS.

A further strain on the classical-model LMS was added when user-generated content (see ‘UGC’ below) came into the learning mix.

The content management capabilities of an LXP like Stream start from a place of CMS power and sophistication that assumes a wide variety of content types, and the ability of learners to generate and upload their own content to the system.


More diversity and complexity in the content handled by a learning system inevitably calls for a more sophisticated and powerful search capability: the learning content accessed via an LXP will not only be of very different types, but might be provided by third parties outside the organization, and might even be hosted on sites external to that organization.   

We are familiar with searching on Google, and almost take for granted the AI and natural language processing (NLP) that goes into delivering quick, relevant research results. At the same time, we typically adjust our expectations downwards when it comes to using a search facility on corporate applications and systems. 

Broadly speaking, search on many such systems has historically been of a much simpler type, designed to deal with structured data (i.e. information organized in a database) or to pick out particular words and phrases. This might be adequate for an LMS running solely elearning modules and pdf documents, however, in a more diverse content environment where a fair amount of the data is unstructured, it is unlikely to return particularly relevant results. This leads to valuable learning resources being at risk to be excluded from the search results.

The LXP brings onboard search up to the current state of the art, with AI-driven technology using natural language processing (NLP) that returns relevant, actionable results, meeting the expectations set by the Janet Clarey quote above for a consumer-grade experience when it comes to search.   

And with increased emphasis on self-directed learning, search becomes ever more critical; being the primary tool with which any search for knowledge commences in the modern age.  

UGC (user-generated content)

UGC in this context refers to the content on a system that is placed there not by the organization running that system, or by third-party content providers, but by learners themselves. It covers a huge range of contributions that can be made by learners, from an article or video that someone might have labored over for hours and even days, to something as simple as a ‘like’, created with a single tap on a mobile phone screen.    

The simplest way to understand what this category encompasses is to think of platforms such as YouTube, Facebook and Wikipedia, whose content is solely provided by users – but also sites like Amazon, Tripadvisor and Glassdoor, where reviews and ratings play an important role.   

UGC can include discussion threads, star-ratings, image-sharing, user-made videos, memes, liking, upvoting, polls, etc. 

The LXP was born in the Web 2.0 world, where it is common practice for users to make a contribution of some kind. ‘Learning by doing’ is an important part of learning: allowing users to practice or in other words, flex that new muscle. Another important aspect of UGC is the ability to transfer valuable knowledge within the organization. You might be able to find a hidden diamond in the rough and be able to capture that contribution to share with other learners. A modern LXP should include the required technology in order to support UGC and the varied benefits that accompany this feature. 

API (application programming interface)

A key phrase in Janet Clarey’s quote above is ‘integrated technologies’. The API is that piece that does the integrating; an interface that allows applications to talk to each other. It’s a very controlled conversation. Think of it as a door you need the right key to get through – but once you are through the door, there are strict rules as to how you can interact. The deal is, if you obey the rules, and make your requests according to the specified format, they will always get a predictable response.

An API allows an application to use the data or functionality of another application, without getting into the wiring of how that other application works. APIs are very powerful and important in modern platforms. We use them all the time without knowing it, for example, when we search for the best deal on a hotel booking or use an ATM.

In the case of an LXP, APIs might be used to draw in data about courses on a third party system. Additionally to give learners access to the functionality of a curation app, like Anders Pink, without leaving the LXP.  


There is one particular, very specialized, type of API that was created specifically for learning systems, known as The Experience API, or xAPI for short. xAPI began life as an update to the widely used SCORM specification, but the ambition of its creators went way beyond that. 

SCORM tracks the elearning courses you take, your quiz scores and whether you complete each course or not. xAPI, on the other hand, makes it possible to record data about a wide range of learning experiences, both offline and online. More than this, it allows this learning data to be collated with business data from other corporate systems. For example, sales training could be compared against individual sales data from a platform like Salesforce, making the link between learning activities and business outcomes.

xAPI activity statements take the form “[actor] [verb] [object]” – e.g. [John] [learned to bake] [cupcakes] – and are recorded in a learning record store (LRS) together with outcome statements – e.g. [John] [won] [star baker on last night’s Great British Bake-off]. Within these statements are additional metadata that gives us even more insight to the user’s activity and further query points.

Since the LXP is fundamentally about the learner experience, being able to accurately record and track those interactions is necessary in the platform.

LRS (learning record store)

The principal function of an LRS is to store xAPI statements from a variety of platforms within an organization’s ecosystem (see above). An LRS typically will sit outside of the LMS or LXP and become the central source of truth.

Traditionally an LMS will limit the data that is able to be exported from the platform, which tends to lock organizations into a solution for a number of years. Whereas an LXP should be compatible with an LRS, allowing data to freely flow in and out of the platform for both reporting and personalization purposes.

Beyond storing data, the LRS allows in-depth learning analytics, since xAPI allows the recording of many more data points than SCORM. Coupled with the ability to correlate with the other sources of data within an ecosystem, L & D teams will have more insight into their learners and the effectiveness of their programs.

Lastly, the LRS being an open system can send data out to a variety of other tools and platforms. This could be to a BI tool for deeper analytics, back to the LXP or LMS for in-platform reporting, badging, certifications or further personalization of learning programs.

Learning Pool believes that LXP and LRS belong together, forming a natural foundation for the modern learner experience.  

AI (artificial intelligence)

AI is a big subject, so let’s just focus on the two main uses made of it within Learning Experience Platforms (apart from search, which is covered above). Providing personalized recommendations and conversational bots, also known as virtual assistants.

We are all familiar with AI-driven recommendations in our lives as consumers. Platforms such as Netflix, Amazon and YouTube all keep a close eye on our wants and viewing habits so they can help us select what we should watch or buy next. Since advertisers (with the use of APIs) are looped into our consumption habits, recommendations follow us round the web as we browse. One might feel like the entire web is one big AI-driven recommendations system. 

The way this works in the case of learning is perhaps more benign.  Learners can record their preferences for what they want to learn on the LXP, but the system can also make inferences based on their behavior on the platform and recommend relevant follow-on learning, for instance, or relevant experiences and resources to complement the learning they are currently undertaking. We can also see the use of AI to make recommendations based on similar users activities and added experiences.

With machine learning these recommendations get more accurate in their relevance over time.

Bots are another feature of our consumer lives, appearing on many websites to help us navigate to what we need via text chat. Some are so clever they make us think we are chatting to a human. Others are merely irritating – it’s all down to how sophisticated the technology behind them is, and how intelligently they have been programmed and designed. 

Bots make use of natural language processing (NLP) to draw inferences from what we contribute in our exchanges with them about what it is we are trying to achieve, and to return relevant suggestions and signposting. NLP is one of the really tough problems in AI, which is why the sophistication of the technology used is such a key determinant of quality conversations. 

However, even within the limits of current technology, bots can do a great deal to help and support learners in answering queries and signposting the resources and experiences they need. The primary aim here is not necessarily to pass the Turing test and fool people that they are receiving human advice, it’s to take friction out of the process of learning, thereby making it more accessible and attractive to learners. 

Next steps

Now that you have an overview of the fundamentals of the technology, you’ll be perfectly equipped to get the most out of our white paper: Powering the Modern Learner Experience: Next-Generation Learning Tools Come of Age. This paper examines the drivers behind the growth in demand for Learning Experience Platforms, and how it should sit alongside the LMS in your decision-making, as well as containing original research on the markets for LXP and LRS.

Get the paper now

business e-learning
learning and development

Got a learning problem to solve?

Get in touch to discover how we can help

CTA background