HackerRank sits in an interesting spot, because it is two things at once, a self study practice platform and the assessment tool that a large number of employers actually use to screen candidates, and both of those are relevant to anyone working toward an AI or data role. As a practice site it is free and organised into domain tracks, so rather than facing an undifferentiated wall of problems you can work through Python, SQL, or problem solving in a reasonably structured sequence, which makes it noticeably friendlier for beginners than jumping straight into interview grinding. The skill certifications are a nice touch, giving you a concrete result you can put on a profile or mention to a recruiter, and while nobody should overstate their value they are a small, honest signal of competence. The reason I tell people to at least create an account is more practical though, because if you are applying for engineering jobs there is a real chance one of your first round screens will be delivered through HackerRank, and being comfortable with its editor and submission flow removes a small but genuine source of interview stress.
The weaknesses are just as real. The problem set is uneven, and once you get past the fundamentals the harder algorithm questions simply are not as well curated or as current as what you find on LeetCode, so serious interview preparation usually means moving on. There is no machine learning, no statistics, and no data science teaching here, so like the other coding practice sites in this catalogue it is a fluency builder rather than a source of AI knowledge. The platform also shows its age in places, with an interface that can feel clunky next to newer tools.
My honest recommendation is to use HackerRank early, to get your Python and SQL to the point where the syntax is automatic and to rehearse the assessment format you are likely to be handed, then move to more specialised resources for both deeper interview practice and the actual AI learning it does not attempt to provide.