Glimpse of my working style: I am driven by research problems, mostly in theory (equations -> proofs -> elegant_code). I aggressively target publication deadlines to structure my contribution efforts, but I consciously try to minimize the influence of those for whom success means recognition and research means publication.
I am happy to collaborate with researchers committed to making an impact (I never hitch-hike papers; I take authorship only for significant contributions). I am always ready to mentor willing students. Email: cs18m20p100002[at]iith.ac.in.
PS: I do not have my research statement submitted for my Google PhD Fellowship, but I can share the structure.

My infrequently-exposed non-academic side: I was born with an interest in Spirituality and Mathematics (in a family of Doctors). I am a sports-lover, a self-taught flute-enthusiast, a huge fan of Disney animation, and have some interests in Creative Arts and Literature. Although I was born and raised in Bihar, I learned almost everything in the Garhwal hills of the Himalayas during my undergraduate years. I envision eradicating child labour in the near future, looking for people to join me.

Some essentials to know before starting a PhD:

• Prioritise a better guide (not the same as an advisor), even if it means compromising a little on your area of interest.
• Do not choose a guide based on how someone presents themselves in lectures.
• Do not choose an advisor based on the courses they teach.
Make both these decisions based on recent lab graduates' experiences. Even if you realize these late, do not worry. Just find (ethical, legal) ways to meet your PhD requirements. There is always a way.
• Do not choose a lab without supportive labmates. Unless you are having a perfect life, you will need some support. Some advisors care (like my father did), but most do not or can not. Your labmates are the ones who will support you.
• The institute's ranking does matter. Higher-ranked institutes have better academic peer groups and collaborations. If you're an ML-theory student in India, IISc is the place to go.
• Never say no when your parents ask you to take/extend a leave.

Disclaimer: I am grateful to my PhD advisor under whom I worked on one of my favorite projects.

On collaborative work: Successful collaborations are determined by the match of working styles rather than a match of skills. PhD students, please be aware of the following possibilities (no way exhaustive):

• Collaborators who treat you with, 'If everything goes well, we will claim full ownership; if anything goes wrong, we will dump it on you'. Tip: Have a clear written discussion of what they expect. Somehow meet that criteria & move on.
• A UG makes a $\delta > 0$ contribution to get his name in & then disappears. Tip: Look for students with integrity & not those with flashy coding skills or high GPAs.
• After taking all the headaches alone, after submission, you get to know there were $> x$ authors. Tip: Confirm the no. of authors beforehand so that the work can be split accordingly.
• A professor has fixed his student's place as your co-author & that student is in no mood to contribute anything. Coping mechanism: Think of this as Gurudakshina & don't get bothered.