I have interest in this area as well.
I wouldn't say that adaptive learning is native in Moodle, but the tools in Moodle and in course build tools outside of Moodle that can work natively within Moodle can be adapted to that use with some strategy in the design stage of the course.
Captivate 9 can build SCORM lessons that can be used in Moodle. You can build in pretests with the results controlling what sections are skipped or complete by the student. The nice thing about this option is that it goes beyond just quiz questions and lets you customize the lesson too. Infosemantic is coming out with some new HTML plug ins that also let you pass custom data to the browser cache to be held and picked up in the next SCORM lesson. I plan on using this to even bypass some of the pre-testing because student level data can be carried forward into the next SCORM.
As far as how to customize within Moodle with tools that already exist within it, we have put the random/pools, hint levels, and redo feature to good use toward a psuedo adaptive learning method. It has taken a lot of work to reorganize all of our questions by type (over 40,000 of them), but it is worth it and would have to be done in a fully adaptive learning environment anyway. Once the question banks are organized by question type, we start to build the quizzes themselves. We label the sections using information labels so the students know what question type they are working with. A student that knows the question type well will get the question correct and move on to the next type. If a student misses, they have two hint levels with point reductions of half at each miss. The second hint where all the points are gone will show a work through for the student to study and they are directed to click try again and then redo to replace that question with a new one of the same type to take what they learned from the walk through and try again for full points. They can keep working at new problems of that type all that they need to until they can get at least one question correct from the pool. Even if they do get a question correct, they can still select the redo for added practice to nail it down. We set a required 80% mastery (activity completion) on practice assignments, so even lazy students that might not take advantage of the redo feature at will find they must do so just to move on. If they do not pass at 80%, they must retake the practice and that retake will be a new random combination of questions from the question sets, so they cannot just copy and paste answers in from their last try. Students are also encouraged to retake the practice assignments as part of their exam study. Even if they passed, the retakes are valuable. We have ours set to highest grade so students are not fearful of retakes possible dropping their score.
Note that at the moment there is a bug in activity completion where teacher score over-rides or hand-scoring does not automatically show the next assignment. There is a work around even if it isn't an elegant one. You just enter the student that got a score over-ride into the exceptions in the next assignment and then they can see it.
There is a feature that let's you show and hide activities by groups. We looked to split out students into groups by overall skill this last year in a course or two and use that to add an additional layer of customizing the course to fit the student. It didn't seem to extend into the grade book though, unless we did something wrong. We found that the grade book expected both levels of the assignment to be done instead of just the one they were assigned. It has caused some confusion. To really make good use of restricted access as an adaptive learning tool, the grade book would have to show only the assigned activity and ignore the alternate.
Good design:
When you set the question pools up, organize by type and difficulty. For instance, in chemistry I have questions of the same type broken out into a formula used directly as being grouped into an easy level, medium would be using that formula but solving for a variable that will require some algebraic shuffling, and hard would require students to realize they need that equation in combination with something else they have learned forming a multi-step problem beyond rote procedural knowledge. All three types would be in the quiz for all student, so it isn't 100% like an adaptive quiz which would have kept lower performing students in the simpler questions, but we have found that it has worked well for us nonetheless. I do like that the lower performing students do get a glimpse of how the next levels up differ so they know what they are aiming toward with growth. The walk through at hint 2 lets them immediately study how it is done so that they can grow with each attempt. They can get extra live teacher help if they are still struggling and student and teacher know exactly what to target when they look at the questions that were missed (thanks to each type and level being labeled).