How Intle gates the quality of AI-generated training

Intle checks generated training for structural and technical defects before delivery, including activity density, visual shape and whether answer keys are parseable. These checks cannot establish that the learning content is true, legally current or accessible to every learner. Review the draft against your source material and test the actual learner experience before publishing. This article describes the implemented checks and their limits.

Four stages, two gate checkpoints

Every generation runs the same pipeline: detection works out what kind of thing you asked for (quiz, module, scenario, compliance and so on), planning turns that into a section-by-section blueprint, an optional structure review lets you approve or adjust that blueprint before any content is written, and generation produces the finished sections. We run the quality gates at two of those stages. The first checkpoint validates the plan before we spend tokens generating from it — a plan that is already the wrong shape can never produce a right-shaped package. The second re-measures the same properties on the real generated output, because a plan can pass and then generation can silently drift away from it: a planned flashcard comes back as a paragraph, a decision loses its response mode. Checking twice is the point.

Structural gates: is this shaped like training?

The first family of gates measures density — the balance of doing versus reading. Each content type carries a minimum interactivity ratio: the proportion of sections that ask the learner to do something rather than just absorb text. The thresholds are tuned per type, because an assessment and a slide deck are not held to the same bar:

• Assessment — 0.8 (four in five sections must be interactive)
• Scenario — 0.7
• Quiz and icebreaker — 0.6
• Compliance, microlearning and interactive session — 0.5
• Module — 0.4
• Presentation — 0.3

Ratio alone is not enough, so the same checkpoint also flags a non-interactive streak of three or more consecutive passive sections (the pattern that makes learners disengage), enforces a visual density floor of roughly 30% of sections carrying a chart, diagram or image, and demands response-mode variety — once a package has four or more questions, it must use at least two answer formats rather than eleven identical multiple-choice stems. There is a pedagogical layer on top: for assessments, modules and compliance content, a Bloom distribution gate fails any package with four or more questions where fewer than 30% reach "apply" or higher, which stops the model settling into a rut of recall-only questions. Some types also have shape requirements — modules and compliance packages must include a flashcard section, compliance must include a comparison table, and briefs that describe a process or an escalation path must include a flowchart. You can see how these thresholds play out in our compliance training examples and across the wider examples gallery.

Content-quality gates: is the content actually usable?

Passing structural gates means the package has the expected shape. A second family checks whether individual pieces are well formed. Alt-text checks reject missing or placeholder descriptions; this is a limited check, not WCAG conformance. Visualisation completeness checks table dimensions and chart data. Flowcharts are parsed through the renderer and checked for disconnected nodes, because invalid diagrams can leave the learner seeing source text instead of a visual.

Other checks limit long copy, flag repeated question stems and flashcard fronts, and validate that a scored item's answer key matches its response mode. For example, a single-choice key must identify an available option and an ordering key must be a full permutation. These checks can catch malformed keys. They cannot prove the marked answer is substantively correct. Review scored items in the assessment workflow.

Blocking versus advisory — not every nit is fatal

Violations are classified as blocking or advisory. Structural failures such as insufficient activity density, an invalid visual or malformed answer key can block delivery. Some copy polish and brief-overlap findings are advisory and can remain for the creator to review. Advisory status does not mean a statement is accurate or suitable for a regulated audience.

Repair attempts and a final decision

The generator can use gate findings to repair a draft and can retry a complete candidate within bounded limits. The current code may attempt more than one candidate; the exact number is an implementation detail and can change. A blocking violation that remains after the allowed attempts stops delivery. An advisory finding can remain for review. A simplified fallback can produce a plainer draft when a richer generation path fails, but this is not a promise that every request succeeds.

What we don't check — yet

No gate establishes that the content is true. The checks are structural, technical or heuristic; there is no factual-accuracy or hallucination gate. They do not establish:

• Whether a stated fact is correct — the model can write a confident, well-structured, wrong sentence and every gate will pass it.
• Whether the answer marked "correct" is actually the right answer — the answer-key gate checks the key is well-formed and parseable, not that it points at the truth.
• Whether a regulatory citation is current or accurate — a compliance module can reference the Money Laundering Regulations 2017, or safeguarding guidance such as Keeping Children Safe in Education, and we will not know if the version has moved on.
• Whether the content genuinely matches your brief — brief alignment is a token-overlap heuristic, not a semantic judgement of coverage or emphasis.

Where the gates fit your workflow

In practice the gates provide a structured, interactive first draft that you then open in the editor, fact-check against your own source of truth, and adjust. The output targets SCORM 2004 4th Edition and SCORM 1.2. Our LMS compatibility page records local package checks and a historical SCORM Cloud reference-platform run, each with dated scope; neither is an accessibility-conformance guarantee or proof that every named LMS works with the current build. If you build regulated content, a qualified reviewer must validate every legal, policy, safety and regulatory statement before use.