I get way too much email. And I hate it. Email is intrusive, addictive, distracting, mentally taxing,
and annoyingly integral to my life. As a researcher, I need long unbroken chunks of "thinking time". Email
smashes those chunks into fragments of wasted time. Even if I'm not checking email, a part of my mind
is thinking about the emails that I haven't responded to. Ugh. A significant number of emails come from students interested in meeting me, and it's a major hassle
to set up individual appointments (which involves even more email.)
You're here because you're a student, and want (no, need) to email me. This document is designed
to answer most of your questions without the need to email. Even if we know each other, it's
probably best to check this document.
The short of it is: I will probably lose your email. (Unless, we have an ongoing conversation.)
As a solution, I hold regular open office hours for all UCSC students. Your best bet is to
come and talk to me face to face, during that time. I will be available at my office, E2-347A.
I have a public Google calendar where you
can see my open office hours. I do have travel and other disruptors that affect my schedule, and I will keep this calendar up to date. Please check the calendar before coming by, just to make sure that I will
But now, let me answer your specific question.
Are you a:
• High school student, or someone about to join college
• UCSC undergraduate
• UCSC Masters student
• UCSC PhD student
• Undergraduate or Masters student from elsewhere
• PhD student from elsewhere
I know. You're surprised at the number of options. Read on.
High school student
Q. Do you have research openings?
A. Sorry, but no.
Q. Wait wait! You don't understand. I have done AP CS, I'm a great coder, a great learner,
and don't need much supervision. Can I get an opening now?
A. No disrespect, but I have a long line of more qualified undergraduate and
graduate students wanting to work for me. I must prioritize UCSC students over others.
Q. Can I ask you about the UCSC curriculum and CSE courses?
A. No. I have designed parts of the curriculum and am quite interested in curriculum
structure. But I am not available to answer general questions about it. You can find this
information online. Or better yet, talk to some current students. They will know more than me.
As far as teaching goes, I teach what my chair tells me to teach, and ignore the rest.
Q. I'm visiting UCSC and would like to schedule a meeting with you.
A. I'm sorry, I don't even have much time for our own students.
Frankly, there's not much you'd learn by talking with me. (I'm not that interesting any more.)
You learn a lot more about UCSC from current students.
Q. If I get this correctly, you're neither going to respond to my emails nor meet with me.
So the only way to interact with you is to get into UCSC?
A. I'm impressed. You're quite perceptive. You should definitely apply to UCSC.
Q. Hi Sesh, how are you?
A. I'm doing well, thanks for asking. Ummm...this may be a generation thing, or an age thing, but I'll let you in
on something important. It's generally not a good idea to refer to your professors by first name. I personally do not
mind at all, but I know of many others who do. There is a formal relationship between an undergraduate and a professor,
and many faculty members prefer to maintain the formality. The safest bet is a Dr. So-and-so, or Prof. So-and-so.
It never hurts to treat college as a professional workplace.
Q. So what about you?
A. As I said, I don't mind being called Sesh. If you prefer to go professional, either Prof. Sesh or Prof. Seshadhri
or Prof. Comandur work. For understanding the options, check out my OAQ.
Q. I'm working on a personal project. Can I get some help or advice?
A. Unfortunately, I'm overstretched by my own projects and students. It is unlikely I can give you much time.
Nonetheless, come and talk to me during office hours, and I will try my best.
Q. Can I know more about the CSE curriculum and what courses to take?
A. Of course you can! Just not from me. I don't know about the course schedule, what will be offered,
who is teaching what, etc. I just teach what my department chair tells me to teach. Talk to the other
undergrads, who will know much more about this than me.
Q. Which instructor should I do <course no.> with?
A. Ask your friends. I don't comment on my colleagues.
Q. I really want your advice on what courses to take, in line with my career goals.
A. Just to be clear, if you want to know what courses to talk to finish your requirements or get a degree,
don't ask me. I don't know! You know better what's on the books, and your peers
know more about what is taught. If you ask me about a course I haven't taught, I would search on Google and read out the first hit.
But wait, you could do that yourself...
But if you want advice on CSE (or math) courses regarding
their intellectual content, I can try to give advice in line with your goals. Although, I will likely repeat
what I'll write here. If you're interested in Machine Learning and Data Science, I highly recommend
taking statistics, linear algebra, and optimization courses. If you're interested in theoretical computer science
(courses like Algorithms and Computational Models), I recommend real analysis, graph theory, combinatorial algorithms,
and any proof heavy course.
Q. Great! When will these amazing courses be offered?
A. Hang on. Let me get my crystal ball. <Getting crystal ball> Nope, it doesn't say anything.
Wait...let me Google it. No wait, why don't you Google it?
Q. Do you have research openings?
A. In general, alas, the answer is probably no. Not that I don't have cool projects, but that I'm overbooked
and do not have the time to supervise. At the very least, you need to have done CSE101.
If you have taken a course with me and come to at least a few office hours during the course
(so I recognize you), come by to my office hours. We can discuss your interests and see what is possible.
I usually do not give our summer projects. I find it too short for me to get anything meaningful. If you wish to work with me,
I will require a long commitment. You must either commit to a senior thesis, or at least two 5 credit independent studies (over two quarters).
You would need to devote at least five hours per week on the project. In any case, come and talk to me.
Q. I took a course with you, and would like to be reader/tutor.
A. Excellent! Check if I'm teaching any courses that you're interested in either grading or tutoring. Before the corresponding quarter starts,
come to office hours. We can discuss details.
Q. Can I get a permission code for <course no.>?
A. Maybe, but emailing me won't help. Come to the first day of class, and I will discuss permission codes.
Q. I'm taking a course with you and couldn't make the normal office hours. Can I stop by your open office hours?
A. Well, I don't want to encourage you to do this regularly. But yes, it is open to all students, so you can.
Be forewarned that I might not prioritize you over others, and I will not give as much time or detail as course office hours.
I often find myself explaining similar things to many students, so it's best you ask questions during normal office hours.
Q. Can I stop by and say hi?
A. Sure! But I might respond and go back to whatever I'm doing. Not being rude, but hey, I have lots of email to delete.
Q. Why is your webpage so bare bones?
A. I find it refreshing to have a screen with just text and (useful?) content. No colorful text to take your eyes away, almost no links
to distract you, just HTML 1.0 text. When was the last time you read a Wikipedia entry top to bottom? The plethora of links
always lead you somewhere else. Here, you have no option but to just read. (Or go away.)
Q. Can you give my advice on my future career, including graduate school?
A. Most certaintly, please come to office hours. I would strongly prefer that you've taken a course with me, so I have some context to give useful advice.
Q. Should I join the MS program at UCSC?
A. In these days of increasing specialization, it's never a bad idea to stay in school just a bit longer to build
your foundations. The faculty like to encourage our undergrads who want an advanced degree. Since it's likely that some
of us know you (and will write letters for you), your odds of getting in are much higher than outside students. If I have
had extensive contact with you during courses, then we can talk about specific details.
Q. Can you write me a letter?
A. I see hundreds of students every year. Unfortunately, simply doing well in a course of mine is not enough
to get a letter. You must have had extensive contact with me. This could be through taking a graduate course with me
and having some technical conversations with me. Or this could be through tutoring my courses, or doing research with me.
If you do satisfy these conditions, come by to office hours to discuss this.
Q. Why don't you (or the other faculty) have any time for us? We're paying to be here!
A. I sympathize with you. (No really, I do.) The answer is complicated. It has to do with how academia
is structured and the incentive structure for academics. The short answer is that research is our primary goal, and it takes
the bulk of our time.
Q. Your office hours don't work for me. Can we schedule another time?
A. I'm sorry, but no. The whole point of office hours is to not deal with scheduling individual appointments. I know, you probably
have a course at the same time, but maybe we can talk another quarter?
Q. No please! This is a life and death situation.
A. Then stop reading and dial 911.
Q. You have not answered <insert unique question>.
A. Then come to office hours and tell me what I missed.
UCSC Masters student
Q. Do you have openings or can you supervise my thesis/project?
A. I usually do not supervise MS students, unless they have done a BS at UCSC (and done some advanced courses with me).
Most of my projects require significant algorithmic and mathematical background, and are unsuitable for 1-2 quarters of work.
Q. Can I tutor/TA/grade your course?
A. For tutoring or being a reader, yes. I am always looking for support for large courses. Please check if I'm teaching a 200+ student course,
and come to office hours before/at the beginning of the quarter. TAing is unlikely. I prefer PhD students for TAing, and the assignment is done centrally
in the department.
Q. Can you be a reader for my thesis?
A. Only if your advisor recommended my name. Come talk to me during office hours.
UCSC PhD student
Before I answer questions, if you are a CSE PhD student, then you have highest priority among students. I am happy to talk
to you about anything, and you are welcome to my office hours to get to know me. I consider it an important duty of mine to
give time to PhD students.
Q. Thank you, Professor.
A. Please call me Sesh. I insist. As a graduate student, we expect you, during the course of your PhD, to grow
into a junior colleague. During research meetings, I want you to challenge me and eventually teach me about your problem.
Therefore, you have the right to address me as a peer.
Q. Do you have openings? I'd like to work with you,
A. Given the length of the answer, this is the topic of a separate essay.
Q. Things aren't working with my advisor. Can you help me out?
A. This is delicate. I cannot take you on as a student, if there is no fit. If I simply take you on
to "save" you right now, we will eventually run into problems. Your best bet is to talk to the graduate
advisor. Suppose you are genuinely interested in my research agenda, and you satisfy my requirements
(as detailed in the previous answer). We can certainly talk more. But be forewarned that I will
also talk to your current advisor to get his/her perspective.
Undergraduate or Masters from outside UCSC
Q. Do you offer summer internships?
Q. I am a prospective PhD student. Do you have openings?
A. Please read my essay on working with me. If you think there will be a match,
please apply to our graduate program. It would help if you mention me by name in your statements, and also give some specifics
about my work. The graduate committee will read the application and forward it to me.
PhD from outside UCSC
Q. Do you have postdoc openings?
A. It's a complicated function of my funding, my current group size, and my potential interests.
If you don't publish in TCS or data mining venues, then there is no fit. Please note that I do not really
work in ML or AI in the traditional sense. I am an algorithms researcher whose work has bearing in these areas.
If we have met at a conference, you are familiar with some of my work, or your advisor knows me, send me an email
with your CV. Just to emphasize, I will not respond if you do
not satisfy these criteria. (Sorry, but I get way too much email.) If you do and I have something, I'll get back to you in a week.
Q. Can I get your advice on a research problem?
A. Unless we have discussed this in person, the answer is no. I am short on time, and prefer to give direction/new ideas
to my own students.
Q. I have a specific doubt about your paper. Can I ask you?
A. Yes please. As a scientist, it is my duty to answer questions about my work. Please ask specific questions.
Such as, "Fig. 3 of <paper> seems to have an error, or cannot be recreated because of..." or "Lemma 2.7 of <paper>
seems to have the following counterexample, or the proof is flawed because of..." It is not possible to answer
generic queries like "can you explain Claim 5.7 of..." or "I don't understand Algorithm 2 of..."