Blog
The Lurno blog
Guides, product notes, and field notes on running learning at scale.
- 01What is an agentic LMS, and the one question that separates themAgentic learning platforms let AI agents do multi-step work by themselves. The difference that decides procurement is whether the agent applies that work, or hands it to a person.
- 02Sana Labs alternatives after the Workday acquisitionWorkday completed its acquisition of Sana in November 2025 and relaunched it inside the suite in March 2026. How to decide at renewal if you don't run Workday.
- 03Docebo alternatives for a mid-market L&D teamHow to run a Docebo evaluation when the renewal is close: reading a quote-only price, scoping implementation, and working out your real per-learner cost.
- 04Moodle alternatives for school groups: cost the operation, not the licenceMoodle is free to licence and not free to run. What hosting, upgrades, plugin compatibility and multi-tenancy actually cost a group of schools.
- 05Canvas alternatives for institutions that have outgrown the academic termCanvas assumes a term, a roster and a transcript. If you run continuing education, client academies or a commercial training arm, here is what to do.
- 06Your course platform breaks when the buyer becomes an organisationKajabi and Teachable are built around one account with one audience. Here is why selling to organisations breaks that model, and how to tell when it has.
- 07Portals, sub-accounts and nested organisations: the multi-tenant question for training providersPortals, sub-accounts and nested organisations are three different things. Which one your platform uses decides what you can sell a client, and at what price.
- 08Saudi Arabia made AI a compulsory subject. What that asks of the technology underneathSaudi Arabia made AI compulsory for more than six million pupils. What that implies for Arabic delivery, assessment, teacher capability and data.
- 09Choosing an LMS that actually works in ArabicAn Arabic translation and a product built for right-to-left are different things. What breaks when direction comes last, and the hour-long test that finds it.
- 10Data residency questions for education buyers in the Middle EastWhere data sits at rest is one question of seven. A vendor-neutral guide to residency, sovereignty and jurisdiction for education buyers in the Middle East.
- 11Running a multi-campus school group on one platformEach campus wants autonomy, the group wants comparable numbers, and both are right. What to share, what to keep local, and why group completion figures lie.
- 12Sub-organisations, branches and groups — and when you need eachA sub-organisation is a child tenant. A branch is an internal tree that roles cascade down. A group is what people are enrolled through. When to use each.
- 13The white-label checklist: what to ask a vendor after the logo uploadWhite-label is not a logo upload. The questions to ask about domains, sending domains, reset links, inheritance and what your page source gives away.
- 14How one company runs an academy for fifteen client organisationsFifteen clients, one academy, one authoring team. What has to be shared, what has to be separated, and what happens when a client wants the course changed.
- 15Eight security questions to ask an LMS vendor, and the follow-up for eachTenant isolation, audit tampering, deletion, the 72-hour clock and sub-processors: eight questions for an LMS vendor, and the follow-up that tests each answer.
- 16Competency frameworks in practice: from standards to evidence to masteryCompetency frameworks die in a spreadsheet the assessment never touches. What mapping, evidence and mastery decisions require, and what a reviewer wants.
- 17Guardian access without oversharingFamilies want visibility, learners deserve proportion, and the school needs one answer it can defend. Why the data category is the unit of consent.
- 18What "AI-native" should mean in an LMS, and how to test itAlmost every learning platform now claims to be AI-native. Four questions — grounding, action, correction and audit — separate the ones that mean it.
- 19Grounding AI course generation in your own documentsUngrounded AI writes plausible material that quietly contradicts your syllabus. How parsing, hybrid retrieval and citations make a draft checkable.
- 20AI grading that instructors trust: drafts, not decisions“The model drafts and the instructor reviews” decays into “the model grades” unless the software stops it. Five conditions that keep the loop real.
- 21An AI tutor that will not leak the answerAsk a general assistant to rephrase an exam question and it often rephrases it into the answer. Why tutor modes belong on the server, not in a system prompt.
- 22Ask your data: natural-language reporting without letting a model write SQLText-to-SQL hands a model your database. The safer shape: it proposes a validated report specification, run under the asking user's own permissions.
- 23Measuring wellbeing without surveilling the people you are measuringYou cannot support people you know nothing about, and monitoring destroys the trust that makes disclosure possible. How to design out of that corner.