2-month process with 8 rounds:
1) online OA (passed) - DSA-like problem based on a real-ish scenario, ~LeetCode medium. Arrays and hashing based.
2) HR screen (passed) - standard half-hour phone screen.
3) Technical screen (passed) - HR contact told me this would be a review of my OA; instead it was a combination behavioral + system design interview. This was the first of multiple completely inaccurate descriptions of what to expect for an upcoming interview. The interviewer himself was great, and obviously went well-enough that I proceeded to the loop.
4) 5-round, half-day interview loop (failed) - each round 45 minutes, with 15 minute breaks. Language-agnostic, use your own IDE. Each had 1-2 behavioral questions on the front end so you really have 30-35 minutes for the actual interview substance:
4a) Coding: DSA. At least, it was told to me by HR that this would be DSA; instead, it was based on a real-scenario prompt, not quite DSA, not quite LLD, not quite L&M: "write an algorithm that places trains on a track given these constraints...", where you needed to draw out requirements, determine what was in/out of scope, and *also* expected to speak to DSA concerns (time and memory complexity, etc.). Other candidates got straightforward DSA prompts, bad luck for me.
4b) Coding: Logical & Maintainable and Problem Solving. Very similar style of prompt to round 4a, except you are expected to speak to the topics in the title of this round rather than DSA topics. If not for the title differences between 4a and 4b, I would have not known that these were evaluating anything different.
Note on 4a and 4b: for a junior-level interview, it was surprising just how silent the interviewers were and how much they expected me to take charge of all conversation; this is more typical of senior-level interviews.
4c) Service Design: a straightforward system design prompt, where you determine requirements, write API endpoints, draw a high level diagram, etc. Unsure why they call this "service" design when it's just system design.
4d) Existing System Review: HR's biggest fail. I was explicitly told this would be "reviewing a new-to-me system," "e.g. simulating a code review". Instead, this was *presenting* a system I have designed and implemented, fielding questions about its scalability, security, etc. Fortunately I happened to be ready for this completely different interview, by sheer luck.
5) Behavioral. 6 STAR questions which were done after 20 minutes, so I filled the rest of the time with questions of my own. Interviewer seemed ready to leave asap/didn't want to be there.
Multiple directly false statements by my HR contact for what to expect, multiple mysterious/concealed requirements for rounds (why call it DSA when it clearly was not?), and the total lack of feedback, drawn out over a 2-month period, made this overall a negative experience. Either I was assigned someone in HR who had no idea what the process actually was, or there is a lot of internal mess at BNSF. Or both. There were some individuals who seemed like people I'd enjoy working with; but that doesn't outweigh the multiple other failures in how BNSF is running this process.