The number of interviews varies from person to person, but most interviewees have 2-4 interviews in all.
The first interview is almost always a phone screen. It is generally a mix of behavioral and technical questions and lasts about a 30 minutes. You may or may not have to write code, but you should be prepared to do so regardless.
The second interview is generally a set of interviews (2+) held on the same day in succession. These tend to be a little bit longer (around 45 minutes each). Most interviewees are generally asked to write code in this set of interviews - it tends to be more technical than behavioral on average but experiences vary. The questions tend to revolve around common data structures and algorithms.
The main crux of behavioral questions is to get to know you beyond what your resume says. Most questions will be open-ended but you are encouraged to showcase honesty, integrity, collaboration, and leadership through past experiences.
While not always warranted, it might be a good idea to think about your past experiences where you have demonstrated these qualities, or past experiences that instilled these qualities in you. Reflecting on the past, gaining insight from it, and improving yourself for the future is a quality of great value.
Sometimes, you might be asked about projects or activities you've stated on your resume. This is also a good avenue to showcase your passion for a field of interest and how you've created an impact with/without technology. You probably won't focus too much on the 'how' (as in, how you implemented a system), but more on the 'why' or 'what' of your projects. You can feel free to briefly mentioned the technology you used, especially if it falls in line with what the team you're interviewing with uses regularly, but that doesn't have to the core focus of the conversation.
As mentioned previously, technical questions vary around common data structures and algorithms.
There are several online platforms and books that can serve as great practice resources for such interviews. It is recommended to not just know technical details but also learn how think out loud through mock interviews with friends.
Apart from these algorithmic-style questions (and, as a matter of fact, complementing your ability to solve those questions), you can try to implement common data structures from the ground up and be familiar with make time-space tradeoffs.