What exactly is the difference between an LMS and an LXP?
A Learning Management System is top-down and compliance-driven: the organisation sets up courses, assigns them and measures whether people have completed them. It works well for mandatory training. A Learning Experience Platform puts the employee at the centre: recommendations based on role and goals, content from multiple sources, peer recommendations, microlearning and self-chosen learning paths. Many organisations run both side by side — an LMS for compliance and an LXP for voluntary, personalised learning. Which route suits you depends on your learning culture and the nature of your content.
Do you use Moodle, TalentLMS or another off-the-shelf package as a foundation?
No, not as a foundation. We build our custom LXPs from the ground up on modern web frameworks (usually Next.js plus a Node or Python backend) with PostgreSQL as the data layer. Off-the-shelf LMS packages such as Moodle or TalentLMS have a data model and plugin architecture designed for compliance learning — working against that structure in an LXP context often costs more than starting from scratch. What we do do: integrate with existing LMS instances if they stay in your landscape for the compliance side.
Which content standards do you support?
SCORM 1.2 and 2004 for classic course packages, xAPI (Tin Can) and cmi5 for modern tracking scenarios, including learning outside the platform, LTI 1.3 for integration with external content providers and H5P for interactive content. A built-in learning record store captures all activity regardless of the source, so you have a single, central view of learning activity. For video, we usually use a dedicated video library (Mux, Cloudflare Stream or an internal S3 environment) so that hosting isn't tied to a third party.
Can we migrate content from our current system?
In most cases, yes. SCORM and xAPI packages are relatively easy to migrate because they are built on standards. Content stored locally in a legacy system (video files, PDFs, hand-made HTML pages) requires a migration script that we draw up case by case. Where possible, we migrate metadata such as skill tags, target-group assignments and learning paths in an automated way; this is often the moment to clean up the taxonomy right away rather than carrying over errors from the old system.
How exactly does the AI personalisation work?
On three layers. First layer: role-based — a new employee in a particular team gets a starting recommendation based on what colleagues in the same role have completed. Second layer: skill gap — the platform compares the skill profile from self-assessment and performance data with the role profile and suggests content to close the gaps. Third layer: behaviour — what someone has already completed, what they have liked or saved, and what their peer group is following. For larger programmes, we add a Q&A bot that answers questions based on your own course material, with source references so employees can verify where answers come from.
Does the platform integrate with our HRIS and Microsoft Teams or Slack?
Yes. For HRIS, we typically build a two-way integration with BambooHR, Workday, SuccessFactors or AFAS via API or SCIM, so that role changes, joiners and leavers, and performance data flow automatically into the LXP — and learning activity and skills progression flow back to HR. For Teams and Slack, we build a bot that handles recommendations, reminders and learning questions directly in the chat, so employees don't need to log in separately. SSO via Azure AD, Okta or Google Workspace is standard.
What determines the cost of a project like this?
The scope (how many modules and what functionality), the depth of integration (how many external systems are involved and in which direction), the presence of AI components such as a Q&A bot, the number of languages and regions, and how much of the content or metadata you want to keep managing yourself. A personalisation layer on top of existing content is a different project from an end-to-end platform with AI and deep HR integrations. After the scoping phase we give you an honest estimate; without that context, any upfront figure would be a guess, and we'd rather not guess.
How long until we can go live?
A well-defined first version — for example, a personalisation layer for one department or a pilot cohort of a few hundred users — can go into production in a few sprints. A broader project with multiple integrations, AI components and a rollout to thousands of employees takes several sprints. We always work in phases: first one target group goes live, we gather feedback and then scale up. We only expand the scope once that first group is running stably on the platform.
What about GDPR, privacy and retention of learning results?
Learning data is personal data and falls under the GDPR. We build role-based access in which only the employee and — depending on your policy — their direct manager can see individual results; aggregated reports for L&D cannot be traced back to individuals. Retention periods are configurable per data category (progress, certificates, sentiment), so you can set how long results are kept after an employee leaves. For AI components, we run models either self-hosted or via Azure OpenAI with EU data residency, and course material never leaves your own tenant.
Do you work with our internal L&D, HR and IT teams?
Almost always. An LXP touches L&D (content and didactics), HR (roles and talent cycle) and IT (identity, security, hosting) all at once. We start with workshops involving all three disciplines, have a dedicated contact person per team throughout the build, and hand over at the end with runbooks, documentation and handover sessions. We build in such a way that you can take over the platform yourself, even if you initially opt for our managed service.