Percipio occupies a category that course review sites tend to ignore, which is learning platforms you did not choose. Somebody in your learning and development team signed an enterprise agreement, a tile appeared on your intranet, and now you have access to several thousand courses you will mostly never open. Reviewing it as though it competes with Coursera misses the point, because it is not really selling to you. It is selling to the person who has to demonstrate that training happened.
Understanding that explains almost every design decision in the product. Start with what it does well, because there is genuine value here. The breadth is real. Skillsoft has been in corporate learning for decades and the catalogue reflects that, with hundreds of learning paths and channels covering technology, leadership, business skills and compliance.
If you need to know roughly what a transformer is before a meeting, or want a working vocabulary for generative AI so you can follow a strategy discussion, there is material here that will do that job in twenty minutes and it costs you nothing. The short module format genuinely fits into a working day in a way that a twelve week specialisation does not. The included book library and practice labs are, in my experience, the most underused and most valuable part of the whole thing. Percipio bundles a large collection of technical books, and reading a proper book through a licence your employer already pays for is a far better use of the platform than working through its own video courses.
Similarly the hands on labs, where they exist, are worth more than the lectures around them. Skillsoft has pushed hard on AI features in recent years, including a conversational practice tool for difficult workplace conversations and AI driven skill mapping across the platform. The coaching simulator is more interesting than I expected and is a sensible use of the technology, letting people rehearse a performance review or a difficult negotiation without a human on the other side. Whether it changes behaviour is another question, but it is a legitimate idea rather than a bolt on.
Now the problems. The AI and machine learning content is wide and thin. You can find courses on Python, machine learning fundamentals, generative AI and data science, and almost all of them stop at an introductory level. There is no path here that takes somebody from nothing to employable in an AI role, and the material does not go deep enough to build real capability in any single area.
Compared with what DeepLearning.AI, Fast.ai or a decent university course will give you, this is a summary of a summary. The quality is also inconsistent in a way that a curated catalogue should not be. Some courses are well made and current. Others were clearly produced years ago, use dated tooling and examples, and survive in the catalogue because nothing forces their removal.
There is no reliable signal telling you which is which until you are five minutes in. That variance is the single biggest practical complaint and it is structural, because a catalogue this large cannot be maintained to a consistent standard. The deeper issue is what the platform is optimised for. Enterprise learning is measured in completions, hours logged and compliance rates, and Percipio is built to produce those numbers efficiently.
Short modules, easy assessments, generous progress tracking. None of that is dishonest, it is simply a different objective from making somebody good at their job, and where the two conflict, the metric wins. The result is a system where it is entirely possible to complete a great deal of AI training and be no more capable than when you started. And the credential is worth nothing externally.
A Percipio completion is a line in an internal system. It is not an AWS certification, not a university certificate, not a portfolio project. If you are trying to move into an AI role, this will not help you do it, and you should not spend your learning hours here expecting it to. My 2.8 is not a judgement on Skillsoft's competence, which is considerable in its actual market.
It is a judgement on whether this is a good place to learn AI, and the answer is that it is a convenient place to get oriented and a bad place to get skilled. The honest advice is straightforward. If it is on your intranet, use it for compliance training you have to do anyway, raid the book library, try the practice labs, and take an introductory AI course or two to get the vocabulary. Then go and learn the actual material somewhere that had to earn your attention rather than your employer's procurement budget.