The process in Boronia was: 1st Phone interview, 2nd: in person interview, 3rd: you are given a task to complete (lasts around 3-4 weeks), and if they like what you've done, they invite you to present it in the office, and then they make an offer.
The first was a phone interview. This was a thorough check of my background in science. My technical skills and my publications. The interviewer was interested to see if I really had the knowledge of all the things mentioned in my CV and also if I have the necessary persistence for "impossible" tasks. The interview lasted ~ 1h. The interviewer was polite and friendly, and I didn't feel uncomfortable in any stage. In the first interview I was also asked about questions on machine learning with regards time series. This is significantly different that standard machine learning techniques, so do some studying before the interview if this is something you don't know (stackoverflow it). It also helps to spend some time reading introductory things in finance (focus on time series analysis). I wasn't asked/tested any particular technical question that made my life difficult.
The second interview was in person, with two interviewers. Again very friendly and very polite. The process was similar (in fact easier) and enjoyable. I was asked how my skills could be transfered in their field. Here helps an itroductory studying in finance. Again, I wasn't asked/tested any particular technical question that made my life difficult.
The third task was a research paper in finance, with their data to reproduce some parts of it. Bottom line: if you don't have finance background, or time series analysis knowledge, you'll probably crash and burn - at least for the particular paper I was given. The computations are pretty easy (not more than a couple of hours in python/matlab), but understanding what equations you need to use for the calculations (many subtle points in the paper) is not entirely evident from the text (or related ones ...). It **requires** finance knowledge (not just time series).
The catch here is that despite the fact the job advertisement was mentioning "finance background not necessary", I was asked in 3 weeks (having also a standard job), not only to reproduce research in finance (with zero knowledge on the topic - stated from the beginning) but also propose new research ideas on improving the paper (otherwise you don't get an offer). And there are papers that are fairly well explanatory, but the one I was given, had several things I couldn't make sense in it.
Overall it was a positive experience and I've learned a lot of things (in finance), but time consuming. I wouldn't propose it for someone who doesn't know finance (i.e. econometrics, time series analysis and basic trading strategies).
Pros:
Friendly interviewers. Enjoyable interviews. You learn a lot of things if you take the task.
Cons:
The task was for people with finance background, so the description of the job advertisement and what they asked was inconsistent. In fact during the interview it was stated explicitly from the head of research that they expect me to contribute in research after the fist 12- 18 months of employment. This was in direct contrast with the task I was given.