Guides for training that stays true
Practical guides on drafting, reviewing and maintaining training from the documents you already have.
How to turn a PDF into a training course (without losing what matters)
A practical, step-by-step method for turning a policy, handbook or guide into a short course people finish, with or without AI.
Training drift: why courses quietly go wrong, and how to measure it
Training drift is the gap between what a course teaches and what its sources now say. Here's how it happens and a simple way to audit it.
A reviewer's checklist for AI-drafted training
AI can draft a course in minutes. Here's a 12-point checklist subject-matter experts can use to review it in under an hour.
SCORM, xAPI, cmi5 and QTI explained: which one does what
Four e-learning standards that are often confused. A plain-language guide to what each one is for and when you need it.
Writing knowledge checks that actually check knowledge
Most quiz questions test recall of trivia. Here's how to write multiple-choice checks that test whether people can apply what they learned.
From SOP to training: making procedures learnable
SOPs are written to be followed, not learned. A guide to turning procedures into training that sticks, safely.
Writing narration scripts for e-learning: a field guide
How to write scripts that sound natural when spoken, with pacing, length and structure that keep learners listening.
Training pull requests: version control for what people learn
Software teams don't edit production by emailing files. Training shouldn't either. The case for diffs, reviewers and releases in L&D.
Completion isn't readiness: measuring whether people are current
Completion rates tell you who clicked through a course once. Here's a better question to ask, and the signals that answer it.
Microlearning from long documents: how to split without losing the thread
Long sources make vague courses. A method for splitting big documents into short, connected lessons.
Turning meeting and webinar transcripts into training
Recorded sessions hold great explanations. How to turn transcripts into structured lessons, and the rights and privacy rules to follow.
Questions to ask any AI training vendor about your data
Before you upload a policy to an AI tool, ask these questions. Our own answers are included.
What is LearningOps? Running training like an operation, not a project
LearningOps applies the habits of software operations to training: sources of truth, change detection, review gates, releases and monitoring.
Source-grounded AI training is more than citations
A citation that looks right isn't proof. Four properties of trustworthy provenance for AI-drafted training, and how to check them.
Why learning needs an intermediate representation
Generating each format separately creates drift between them. A typed course model compiled to every output keeps them consistent.
The three-way merge for learning content
When a source changes after people have edited an AI draft, you need to update the facts without losing the human craft. Here's how a three-way merge does it.
How to run a training-maintenance pilot that proves something
Start narrow, prove the full loop, then expand. A six-step guide to piloting AI-assisted training maintenance with real evidence.
Migrating a legacy course library without migrating its mistakes
Moving courses to a new platform is the best moment to find what's out of date. A guide to turning migration into a drift audit.
Calculating the ROI of training operations, honestly
Most AI ROI claims count only drafting speed. A defensible model separates drafting, review and maintenance, and measures each against a baseline.
Put a guide into practice
Free during early access · No account needed · We don't keep your files