YoE: 5 years in top 20 Fortune
PhD form top 20 USA University
Preparation 400 LC mostly medium and hard

Special thanks to LC community, this would not have been possible without countless posts and guidance from the community

Both positions were ML SWE

Rounds at FB: 1 ML/ 1 Syst design / 2 coding / 1 behavioral
The codings:

Not sharing due to NDA. They were all in the top 40 FB qqs in leetcode...

All of the coding I had a solution after the first 5 minutes, started coding afterwards, though I tried to do alot of practice interviews, I still got nervous and delivered buggy code... That's the way I am... I get nervous and things seem to run out of my memory and mind....

They got back to me after a week, I had gone for an E5 but got an E4 stating that my Syst design was not so good...

Rounds at Google: 2 ML/ 2 Coding / 1 Googlyness:
Coding:

Both in the first 40 qq for Google in LC.

Again the nervous problem.... Had working solution in 5 minutes, could not generate "bug free" code... Though I did code it up but the interviewer had to point out the bugs...

They said it is mostly positive but need to find a team so it will make my case stronger in the HC... I got this qq: How long will it take to find a team? Do u have to commit to the team u talked to before going to HC? Or can u change your mind afterwards? Like do u get a team match after the HC or is it before only? I am a bit confused... Like I want to know if I have the offer before it takes too long...

PS. Here an update 1/5/2022:
I did find a team at Google, it took me some time to find one... Though my packet went to the HC and got rejected. After some negotiations I accepted my FB offer...

The ML rounds had the same structure for both FB and Google:

1. ML knowledge : Know one thing in your ML expertise best, for me, it was recommendation systems, I knew it inside out, from Collaborative Filtering, to Wide and Deep NNs, Content based filtering, the Datasets I used and the scope of the problem. Adding here cuz did not have time to reply individual comments... I received alot of help from edu**.io ML course and inte**.io Mock interviews, spent about $1k in interviews but I think it paid off at the end...

2. Previous projects that you performed before on ML topic, related to 1

My experience was very good for FB and awful for Google. Not because I got rejected at the end, but I felt like they did not value anything in me as a candidate... Like one interviewer at Google, the guy who was supposed to talk about my ML experience, asked about the meaning of my name and literally we talked about it for half of the interview time and finally it was one of my No Hire ones.. I mean he did not bother to ask me about my experience... Having a PhD and 5 years of Industry experience did not pique his interest to talk about my ML experience but my simple first name did....

On the other hand, the recruiter had told me finding a group at Google would help my packet... Though I eventually did find a team at Google that would support me, but the process was a Chicken and Egg problem: I was not allowed to see the internal postings unless I joined and I could not join unless I found a group... Only the recruiter could see the job postings and find a group... Meanwhile, the FB offer had come through and I had to make them wait for a couple of weeks...

On the other hand, for FB, I talked to 5 different managers as opposed to 2 at Google and that was not even required... I could not see the internal groups and it was not required to find a group either... Seemed pretty fair and square...

Having all these said, even if I were given a job offer at Google, I liked FB better cuz the groups that I talked to were better.... At Google, once that one group liked me, I wasn't given a chance to talk to anyone else...