Education
Software engineering for E-Learning
E-learning is not a content-management problem with a play button on top. The hard part is that learners drop out, courses go stale, and a platform that scores well on a feature list still fails if nobody finishes the course. The standards (SCORM, xAPI, LTI), are where interoperability lives or dies, and video and content delivery at scale is a real engineering problem. We build for engagement, completion and content that survives contact with actual learners, not for the demo.
Why the domain matters
E-learning is about the learning experience (how content is authored, delivered, experienced and completed), and that is a genuinely different problem from running an institution. The centre of gravity is the learning platform: a learning management system that structures courses, tracks progress and issues certificates, or a learning experience platform that leans harder into discovery, personalisation and engagement. Around it sit authoring tools, assessment and quizzing, video and content delivery, and the interoperability standards that let content and platforms talk to one another. Success here is not measured in features shipped; it is measured in whether learners engage, progress and actually complete, because a course nobody finishes has taught nobody anything however polished it looked.
The uncomfortable truth of the sector is that engagement and completion are the hard problem, and most e-learning underdelivers on both. Learners abandon courses, motivation decays, content ages, and the analytics that were supposed to prove impact often just prove that people logged in. Whether the audience is employees working through mandatory compliance training, professionals pursuing certification, or consumers on a subscription edtech app, the same gravity applies: attention is scarce, drop-off is the default, and the platform earns its keep by keeping people learning. The engineering that matters (interoperability, video delivery, assessment integrity, accessibility), all serves that end, and a platform that treats content as static files to be served has misunderstood the assignment.
We are a senior-led team and we operate what we build, so we start from the learner and the content lifecycle rather than from a feature checklist. The failure mode in e-learning is the platform that ticks every box in a procurement spreadsheet and still sees courses abandoned halfway, because engagement was assumed rather than designed, because the standards were half-implemented and content broke, or because video buffered and quizzes felt like a chore. We would rather build the unglamorous interoperability, the resilient content delivery and the assessment that people trust than the feature nobody uses. This is deliberately distinct from institutional education software, which is about student records, statutory returns and safeguarding, here the subject is the learning itself.
The challenges in e-learning
Engagement and completion are the real problem
The defining failure of e-learning is drop-off, learners start courses and do not finish, motivation decays, and the platform quietly becomes a place people log into rather than learn in. This is not solved by adding features; it is designed for, through pacing, feedback, relevance and honest analytics that measure learning rather than logins. A platform that assumes engagement instead of building for it will score well in a demo and fail in the field, which is exactly where most e-learning underdelivers.
Interoperability standards that are where content lives or dies
SCORM, xAPI (Tin Can) and LTI are the standards that let content and platforms interoperate, packaging courses, tracking learner activity, and embedding tools into a host environment. They are also where a great deal of e-learning quietly breaks: a half-implemented SCORM runtime that loses progress, xAPI statements that never reach the record store, an LTI integration that fails to launch. Getting these standards genuinely right is unglamorous, detailed engineering, and underestimating them is a common way an e-learning project ships content that does not track.
Content that ages and authoring that has to keep up
Courses go stale, facts change, regulations move, screenshots date, and content authored once is rarely maintained on the cadence it needs. The authoring experience has to make updating content genuinely easy for the people who own it, or the platform fills with material that is subtly wrong and slowly erodes trust. Treating content as a one-time upload rather than a living asset with a lifecycle is how a promising platform decays into a library nobody trusts.
Video and content delivery at scale is a real engineering problem
Video is central to modern e-learning and it is genuinely hard to deliver well: encoding for many devices and connections, adaptive streaming that does not buffer, resilient delivery across a distributed audience, and doing it at a cost that works. A course carried by video that stalls, fails on mobile, or is expensive to serve undermines the learning directly. This is real infrastructure work, not a matter of dropping files behind a play button, and it deserves to be resourced as such.
Assessment that is fair, meaningful and hard to game
Quizzes and assessments have to actually measure learning and resist casual gaming, especially where they gate certification that carries weight. That means thoughtful item design, integrity considerations, and feedback that teaches rather than just scores. An assessment that is trivial to pass without learning, or that frustrates genuine learners with brittle marking, damages the credibility of the whole course, so assessment design is a first-class concern, not a form bolted onto the end of a module.
Accessibility for a genuinely diverse learner base
Learners are diverse in ability, device, connection and context, and accessibility is both a legal expectation and a core inclusion issue, captions and transcripts for video, keyboard and screen-reader support, sensible contrast, content that works on a phone on a poor connection. E-learning that is inaccessible simply excludes people from learning, so we engineer accessibility into the content and the platform from the start rather than treating it as a remediation task after a course is already live.
What we build for e-learning
The systems this sector most often needs, built by engineers who understand the domain, not just the code.
LMS and LXP platforms
Learning management and learning experience platforms built around engagement and completion rather than a feature list, structured courses, progress tracking and certification where an LMS fits, and discovery, personalisation and pathways where an LXP is the better shape. We build for the learner’s journey through content, with honest analytics that measure learning rather than logins, because the platform earns its keep by keeping people learning, not by the length of its feature sheet.
Course authoring tools
Authoring experiences that make creating and (crucially), maintaining content genuinely manageable for the people who own it, so courses stay current instead of ageing into material that is subtly wrong. We treat content as a living asset with a lifecycle, building authoring that supports updates, versioning and review, because a platform is only as good as the freshness of what is inside it, and stale content erodes trust faster than any missing feature.
Standards and interoperability (SCORM, xAPI, LTI)
Correct, resilient implementation of the standards that make e-learning interoperable. SCORM packaging and runtime that reliably tracks progress, xAPI statements that actually reach a learning record store, and LTI integrations that launch and pass data cleanly into a host environment. This is detailed, unglamorous engineering that we take seriously, because it is precisely where a lot of e-learning quietly breaks and content stops tracking.
Assessment and quizzing
Assessment that measures learning and holds up where it gates certification, thoughtful item types, integrity considerations, and feedback that teaches rather than merely scores. We build assessment as a first-class part of the learning experience, resistant to casual gaming and fair to genuine learners, because a trivial or brittle assessment undermines the credibility of the whole course and the certificate at the end of it.
Video and content delivery at scale
The infrastructure to deliver video and rich content reliably to a distributed audience, encoding for many devices, adaptive streaming that does not buffer, resilient delivery, and cost that works at scale. We treat this as the real engineering problem it is, because a course carried by video that stalls or fails on mobile undermines the learning directly, and getting delivery right is what lets content-heavy courses actually work in the field.
Certification and credentialing
Issuing, tracking and verifying certificates and credentials in a way that carries meaning, tied to genuine completion and assessment, verifiable, and integrated with the learner record. For corporate compliance training this is an auditable record of who completed what and when; for consumer and professional learning it is a credential the learner can trust and share. Either way we build it so the certificate reflects real learning rather than mere attendance.
Where we help
A corporate training platform where completion actually matters
An organisation rolling out compliance and skills training needs people to genuinely complete courses and needs an auditable record of who did. The real work is engagement and honest tracking: pacing and feedback that get people through, SCORM or xAPI implemented so progress is never lost, and certification tied to real completion. The measurable win is completion rates that reflect learning and a defensible record for audit, not a dashboard that merely proves people logged in.
A consumer edtech app fighting subscription churn
A subscription learning app lives or dies on retention, learners who disengage cancel. We build for the engagement problem directly: pathways and personalisation that keep the next step compelling, video that plays smoothly on a phone on a patchy connection, and assessment that feels like progress rather than a chore. The point is to keep learners learning week after week, because in consumer edtech engagement is not a nice-to-have metric, it is the business.
A content library that stays current instead of rotting
A training team’s course library is slowly filling with material that is subtly out of date, and nobody has an easy way to keep it fresh. We build authoring and content-lifecycle tooling that makes updating, versioning and reviewing content genuinely manageable for the owners, so courses stay accurate. Done well the library stays trustworthy; treated as a one-time upload, it decays into content people quietly learn to distrust.
Reliable video delivery for a content-heavy course at scale
A course built around video needs to play smoothly for a large, distributed audience across devices and connections without a runaway bill. The work is real infrastructure (encoding, adaptive streaming, resilient delivery and sensible cost), because a lesson that buffers or fails on mobile undermines the learning directly. We build the delivery layer as the engineering problem it is, so the content actually reaches learners in a state they can learn from.
How we build for e-learning
We start from the learner and the content lifecycle, not the feature list. Before designing anything we want to understand who is learning, why they drop off, how the content is created and kept current, and what completion genuinely means for this audience. In e-learning the constraint is almost always engagement and content freshness, so we design for the learner who will abandon the course the moment it stops being worth their attention, not for the procurement spreadsheet that rewards a long feature sheet.
We treat the interoperability standards and content delivery as core engineering, resourced accordingly. SCORM, xAPI and LTI are where a great deal of e-learning quietly breaks, and video delivery at scale is a real infrastructure problem, so we plan for the runtime edge cases, the statements that must reach the record store, the launches that must not fail, and the streaming that must not buffer: from day one, rather than discovering them when content that looked fine in the demo stops tracking in the field.
We are blunt about the difference between features and learning. Adding capabilities is easy; building a platform where people actually finish courses is not. We would rather invest in engagement, resilient delivery, trustworthy assessment and content that stays fresh than in a longer feature list that scores well in procurement and still sees courses abandoned halfway. If an idea adds surface area without helping anyone learn, we will say so rather than build it to pad the sheet.
Because we operate what we build, the people designing an xAPI pipeline or a video-delivery layer are the ones called when statements go missing or a stream stalls for a cohort. That concentrates the mind on the failure modes that actually matter in e-learning. The progress that must not be lost, the video that must play on a poor connection, the certificate that must mean something, and keeps us honest about trade-offs up front rather than discovering them in production while a course is live.
Standards and compliance
E-learning is governed more by interoperability standards than by heavy sector regulation, and those standards are where a lot of the real discipline sits. SCORM defines how course content is packaged and how it reports progress to a platform; xAPI (Tin Can) captures richer learning activity as statements sent to a learning record store; LTI lets tools and content embed cleanly into a host learning environment. Implementing these correctly and resiliently is what makes content portable and trackable, and half-implementing them is how progress silently fails to record, so we treat conformance to the standards as core engineering, not a badge.
Accessibility is the compliance dimension that reaches into every course. Learners are diverse in ability, device and context, and accessible design: captions and transcripts, keyboard and screen-reader support, sensible contrast, mobile and low-bandwidth resilience, is both an expectation and a genuine inclusion issue. Where a platform serves a public-sector body it may fall under the public-sector accessibility regulations, and in any setting inaccessible learning simply excludes people, so we engineer accessibility into content and platform from the start.
Where certification carries weight (compliance training, professional credentials, regulated competencies), the record of who completed what, when, and to what standard becomes something that has to be accurate, auditable and verifiable. We build certification and credentialing so it reflects genuine completion and assessment rather than mere attendance, because a certificate that does not mean anything is worse than none where a regulator or an employer relies on it.
Data protection under UK GDPR applies to learner data, which can include children where an edtech product serves under-18s and therefore deserves particular care. We build for lawful basis, minimisation, retention and access from the start, but we are engineers, not your legal, compliance or accessibility advisers. The obligations around accessibility, data protection and any regulated certification carry real weight, and sign-off rests with your own functions. Our job is to build learning software that meets the standards and captures what those obligations require, alongside the people accountable for them.
Integration
The interoperability standards are the first integration surface, and the one that most defines whether e-learning works. SCORM and xAPI govern how content packages report progress and activity back to a platform and a learning record store, and LTI governs how tools and content launch inside a host learning environment. We build these connections to conform genuinely to the standards and to be resilient to their edge cases: the lost SCORM session, the xAPI statement that must not be dropped, the LTI launch that must not fail, because this is exactly where content stops tracking when it is done carelessly.
Content and video infrastructure is the second. Delivering rich media at scale means integrating encoding and streaming, content delivery, and storage in a way that plays smoothly across devices and connections without a runaway cost. For content-heavy courses this delivery layer is not a detail behind a play button; it is core infrastructure that determines whether the learning actually reaches people in a usable state, so we engineer and integrate it deliberately.
In corporate settings, learning platforms rarely stand alone. They connect to identity and single-sign-on, to HR and talent systems that assign training and consume completion data, and sometimes to the broader people or compliance stack. We integrate so that assignments flow in and completion and competency data flow back cleanly, because in an organisation the value of training is realised when its record connects to the systems that act on it rather than sitting in an island.
The recurring engineering task across all of this is keeping learner activity and completion consistent as it crosses authoring tools, the platform, the record store, video infrastructure and any host or HR systems that were never designed to agree. We treat that reconciliation (a single, trustworthy picture of what each learner has done), as core work, because analytics and certification are only as good as the consistency of the activity data underneath them.
Security and data protection
Learning platforms hold a meaningful store of personal data (learner identities, progress and activity, assessment results, and certification records), and where a product serves children or young people that data deserves particular care. We build with access controls scoped to genuine need, encryption in transit and at rest, and retention limited to what a lawful basis actually supports, because learner activity and results are more sensitive than they first appear and should not accumulate loosely across the platform.
Assessment integrity is a security concern in its own right, especially where results gate certification that carries weight. Protecting the fairness of an assessment (resisting casual gaming, safeguarding item banks, and keeping results trustworthy), is part of keeping the credential meaningful. We design assessment and its data with that integrity in mind, because a certificate that can be trivially obtained without learning undermines the very thing the platform exists to provide.
Video and content delivery carry their own considerations: protecting content where it has commercial value, controlling access so paid or restricted material is not freely lifted, and doing so without degrading the experience for legitimate learners. We build content protection proportionate to its real value rather than bolting on friction that punishes honest users, keeping the balance between protecting the asset and serving the learner.
Because we operate what we build, security here is not a report handed over at the end. We instrument for the access and activity patterns that indicate a problem, keep an audit trail that a data-subject request or an assessment-integrity question would need, and treat the ability to reconstruct what happened to a learner’s record, particularly where that learner is a child or the record backs a certificate someone relies on, as part of the deliverable rather than something found missing after an incident.
What changes
Learners who actually engage and complete
Because we design for engagement and completion rather than a feature list (pacing, feedback, relevance and honest analytics that measure learning rather than logins), courses get finished and the platform keeps people learning, which is the only outcome that means the content has done its job.
Content and progress that reliably tracks and stays fresh
By implementing SCORM, xAPI and LTI genuinely and building content as a living asset with a real authoring lifecycle, progress records reliably and courses stay current, so you avoid the two quiet failures of e-learning: content that stops tracking, and a library that ages into material nobody trusts.
Video and delivery that hold up at scale
Because we treat content delivery as the real infrastructure problem it is (encoding, adaptive streaming, resilience and sensible cost), video plays smoothly for a distributed audience across devices and connections, so content-heavy courses actually reach learners in a state they can learn from rather than buffering them out.
What we build for e-learning
From a first platform to modernising what you already run. The disciplines this sector draws on most.
How we deliver
- 01
Discover
We map the system, the constraints and the business it serves, including the parts nobody documented.
Architecture brief
- 02
Architect
Decisions get made, written down and defended before a line of production code exists.
Decision records
- 03
Build
Short cycles against working software. You see progress in the product, not in a status deck.
Shipping increments
- 04
Operate
Monitoring, incident response and iteration. The system is alive, so the engagement is too.
Runbooks & SLOs
Building something for e-learning?
Tell us the problem and the constraints you are working under. A senior engineer will give you a straight view on what it would take, and say so plainly if we are not the right team for it.
Technologies we work in
Chosen per problem, not per fashion. A selection of the stack we most often reach for.
Why teams in e-learning choose us
We build for engagement and completion, not the feature sheet
The defining failure of e-learning is that people do not finish, and features do not fix it. We start from the learner who will abandon a course the moment it stops being worth their attention and design for pacing, feedback and honest analytics, because a platform that scores well in procurement and still sees courses dropped halfway has delivered nothing.
We take the standards and delivery seriously
SCORM, xAPI and LTI are where e-learning quietly breaks, and video delivery at scale is a genuine infrastructure problem. We resource both as core engineering and build them resilient to their edge cases, rather than half-implementing a standard or dropping files behind a play button, which is what separates content that tracks and plays from content that silently fails in the field.
We treat content as a living asset, not a one-time upload
Courses rot when authoring is an afterthought. We build the authoring and content-lifecycle tooling that lets owners keep material genuinely current, because a platform is only as good as the freshness of what is inside it, and stale content erodes learner trust faster than any missing feature ever would.
We operate what we build
The people who design an xAPI pipeline or a video-delivery layer are the ones paged when statements go missing or a stream stalls for a cohort mid-course. That keeps us focused on the failure modes that matter in e-learning, progress that must not be lost, video that must play on a poor connection, certificates that must mean something, and honest about trade-offs up front, not discovering them while a course is live.
Related sectors
Part of Education. Adjacent sectors we also know.
Common questions
How is e-learning different from your education work?
They solve different problems and we keep them distinct. Our education work is about the institution: the accurate student record, admissions, timetabling, statutory returns to the DfE, safeguarding, and the administrative and compliance duties a school, college or university carries. E-learning is about the learning itself: how content is authored, delivered, experienced and completed, and how you keep learners engaged. A platform for delivering courses at scale is a fundamentally different build from a system of record that files a census, and conflating the two is how projects lose focus; if you need both, we treat them as the separate disciplines they are.
Why do so many e-learning platforms fail to keep learners engaged?
Because engagement is the hard problem and most platforms assume it rather than design for it. Learners drop off by default (attention is scarce and motivation decays), so a platform that ticks every feature-list box still sees courses abandoned halfway if pacing, feedback, relevance and honest analytics were not built in. We start from that reality, design for completion rather than logins, and measure learning rather than mere activity. We will also tell you plainly when a requested feature adds surface area without helping anyone actually finish a course, rather than building it to pad the procurement sheet.
Do you support SCORM, xAPI and LTI?
Yes, and we treat them as core engineering rather than a compatibility badge, because they are exactly where a lot of e-learning quietly breaks. SCORM packaging and runtime have to track progress reliably without losing a session; xAPI statements have to actually reach the learning record store; LTI integrations have to launch and pass data cleanly into a host environment. We implement these to conform genuinely to the standards and to be resilient to their edge cases, because content that looks fine in a demo but stops tracking in the field is one of the most common and damaging failures in this sector.
Can you handle video and content delivery at scale?
Yes, and we treat it as the real infrastructure problem it is rather than files behind a play button. Delivering video well means encoding for many devices and connections, adaptive streaming that does not buffer, resilient delivery to a distributed audience, and a cost that works at scale. A course carried by video that stalls or fails on mobile undermines the learning directly, so we engineer the delivery layer deliberately, so content-heavy courses actually reach learners in a state they can learn from rather than buffering them out of the lesson.
Is this for corporate training or consumer edtech?
Both, and the same fundamental gravity applies, attention is scarce, drop-off is the default, and the platform earns its keep by keeping people learning, but the emphasis differs. Corporate training leans on auditable completion, integration with HR and identity systems, and certification that stands up for compliance. Consumer edtech leans on retention, because a disengaged learner cancels, so personalisation, pathways and a compelling next step matter directly to the business. We build for whichever you are, and we are honest that the engineering of engagement, standards and delivery is shared even where the goals and the metrics differ.
Building for e-learning?
Tell us what the system has to do and what it cannot get wrong. A senior engineer reads it, and if e-learning is not a domain we know well enough to be useful in, we will say so rather than learn it on your budget.
- 01A senior engineer reads it. Not a form queue, and not an account manager.
- 02We reply either with questions or with a straight answer that we are not the right fit.
- 03If it looks like a fit, a technical call with the person who would actually run the delivery.
- 04Then scope, effort and risk in writing, before anyone signs anything.