There is no limitation to the type of learning content we can process. If it has human-readable text, we can analyse it and tag it.

Generic skill models are no better than a pick-n-mix. You need a skill model tailored to the world your organisation lives in, its personality traits and its technologies.

Solve your tagging problem. Unique algorithms can generate their own training data and then tag all of your content to a better degree of accuracy than a human.