The first step in the process is an OA, which includes 5 questions: 3 multiple-choice and 2 coding questions in Python. If you pass that, you'll have a technical screen about APIs and data manipulation, which is about 40 - 50 minutes long. I didn't do well on the screen, so I was rejected shortly after. Something I would like to point out, though, is that I was told by my recruiter that the OA would be 2 hrs long and would contain 3 questions. In reality, the OA was a 75 min test that had 5 questions instead. Also, I was told that the screen would have behavioral questions, but I wasn't asked any behavioral questions. Now, this may have just been a series of mistakes, but I think this could indicate that there may be some issues, particularly miscommunication or misalignment, in the company, especially during the interview process, so that's something to keep in mind. Also, I made the mistake of preparing for the technical screen too narrowly, by only focusing on one aspect of data manipulation, when I could be tested on any example of that concept. I practiced calling APIs and processing the data (think SQL queries, but in Python code using JSON data) and even creating APIs that did the same thing to the data. However, what I got was very different than what I studied for. I got a question that, while it was technically an example of data manipulation, was more aligned with system design concepts that I didn't study at all for this interview. Thus, I struggled to explain my approach to the problem and was rejected. I know this is general advice, but for future interviews, please study all the concepts in full; don't just study the aspects of that concept that are widely tested. Study and prepare for any valid example of the given concept. If I had done that, I probably would have explained it very well and done the interview smoothly. Also, be prepared to set up the entire API during the technical screen with either Flask or FastAPI, iirc, as you will have to make the API service from scratch (no supporting code). They will also ask you to run that API and call it to test it.