This fall, I will join the Computer Science PhD program at the University of Washington. Getting there took more than a year of researching programs, writing essays, interviewing, and eventually choosing between several options I was genuinely excited about. Because my interests span computational biology, machine learning, and biological modeling, even deciding which departments to apply to became part of the process.

Coming from the Israeli academic system, I had to learn how US programs admit students while also figuring out where I wanted to do research. This post is about what that year looked like: what surprised me, what I misunderstood, and what I would do differently. It is not a comprehensive guide or a formula for admission, only the account I would have wanted when I began. Much of what I learned came from current PhD students, professors, friends, and recent applicants who explained the system, read drafts, and prepared me for interviews.

The formal application season mostly took place in the fall, but the preparation began much earlier. Mine looked roughly like this:

  1. SpringBuilt a program list and audited international-applicant requirements.
  2. SummerAsked for advice, narrowed the list.
  3. Early fallDrafted my statement, updated my CV, and contacted recommenders.
  4. Late fallTailored materials, navigated application portals, and submitted.
  5. WinterInterviews, rejections, silence, and far too much email refreshing.
  6. SpringOffers, visit days, more conversations, and the final decision.

I created a spreadsheet to track deadlines, requirements, professors, letters, fees, and essays. Starting in spring or early summer gave me about six months to handle testing and transcript requirements before the essays took over my life. The research experience and relationships with recommenders behind those applications had taken years.

I approached the process from the Israeli academic system, where many conventions work quite differently. The first surprise was the scale: admission rates at the US programs I was considering were low enough that applying to only one or two places, even with excellent research fit, would have been a significant gamble. I therefore built a much broader list than many Israelis expected.

This difference is not because PhD programs in Israel are worse or less selective. Israeli admissions are often more personal and PI-driven. In many fields, a prospective student can identify a lab, speak directly with the PI, and develop a mutual understanding before the formal university process. If the PI wants to supervise the student and the practical details work out, admission can feel relatively direct.

In the US, I encountered two broad models. Some programs admit students through a departmental committee, with students choosing or confirming an advisor later. Others admit directly into a particular lab, closer to the Israeli model. In committee-based programs, a connection with one professor did not determine the outcome because class size, funding, and advisor availability also mattered. I had to separate my interest in working with a professor from my chances of joining their program. That was part of why my list grew so long.

The degree structure was another source of confusion. In Israel, students commonly complete a research master’s and then begin a PhD with a specific advisor. In the US, many enter a longer PhD after a bachelor’s degree, with coursework, rotations, and sometimes a master’s built in. My master’s gave me valuable research experience, but it was neither a universal prerequisite nor a shortcut through the US degree.

Funding was another part of understanding what I was applying for. The programs I considered generally waived tuition and paid a stipend through the department, an advisor’s grant, teaching, or a fellowship. I learned to check how many years were guaranteed, whether summers and health insurance were covered, and what the stipend could buy locally. Thinking of the PhD as both training and a workplace clarified why advisor fit, lab culture, and location mattered so much.

Alongside the research application, I had to track requirements that varied between universities. I took TOEFL early rather than risk discovering too late that I needed it; some programs required it, while others granted waivers based on previous education. None on my list required the GRE. Checking each department’s rules was more useful than relying on general advice.

Transcripts raised another set of questions: whether a program accepted an unofficial upload, required an English version, wanted an explanation of the Israeli grading scale, or asked for a credential evaluation. My rule was not to convert my average to a US GPA or reinterpret anything unless explicitly requested.

Application fees, fee-waiver rules, separate recommendation portals, and deadlines in US time zones also accumulated quickly.

Although the problems I care about are biological, I eventually realized that my heart is in computer science. With a computer science degree, industry work as a Production Engineer at Meta, and a master’s in bioinformatics, I was somewhere between fields. I applied to computer science, bioinformatics, and interdisciplinary programs without knowing which academic home fit best.

The same advisor may supervise students from several departments, and the daily research can be nearly identical. The differences often lie in coursework, culture, funding, and community, so I asked about rotations, cross-department advising, funding, and where students with similar interests actually spent their time.

The answer became clearer as I revised my statement and narrowed my lab preferences. I wanted to develop computational ideas and methods for understanding biology, which made computer science feel like the right academic home.

I also realized that I wasn’t really choosing universities; I was choosing advisors. Rankings mattered much less than I expected. I cared more about questions like:

  • Would I enjoy discussing research with this advisor every week?
  • Could I imagine spending several years working on these problems?
  • Did the department encourage the kind of work I wanted to pursue?

Beyond papers, I looked at how PIs presented their labs. Did the website showcase students and their work or only the PI? Where did former students go? Small details also made the people behind the research feel human: one PI had a creative-writing section; another lab had a page for “lab pets.”

By the time I finalized my list, my spreadsheet was organized around professors rather than institutions. I still identified several possible advisors at each program because I wanted to remain excited about it even if my first choice was unavailable.

I applied only to places I could imagine attending. With each application taking time, money, and attention, I did not want to spend those on a place whose offer I already knew I would decline.

I expected cold emails to matter more than they did. Most went unanswered. Some professors said they were not taking students, which helped me filter my list; others essentially replied, “Apply, and we will see.” In committee-based programs, these responses did not seem to determine the outcome, and I stopped treating silence as a judgment on my application.

Existing connections led to more substantive conversations, especially in direct-admission programs where a professor may need to support the application and have funding. When writing without a connection, I mentioned a talk, paper, or other concrete reason their research related to mine. The useful question was not merely whether a lab had an opening, but why I was writing to that person.

Most programs asked for transcripts, three recommendation letters, a statement of purpose, and a CV; some added a personal essay. That list sounds straightforward. Figuring out what belonged in a statement of purpose, and how to write it for readers in the US system, took me many revisions. Even once I had a draft, different prompts and word limits meant I had to revisit it for every application.

My first statement was a record of what I had done. I initially treated it too much like a chronological autobiography. Over many revisions, I shifted the focus to what I wanted to study next, why it mattered, and how my previous work had prepared me.

I also had to learn American academic conventions that were not intuitive to me. One small example was learning that calling someone’s work merely “interesting” could read as faint praise. I became more specific about what I admired and why it connected to my questions.

Feedback from people familiar with the US system helped me see both what was missing from the draft and what readers might expect me to spell out. I wish I had understood those expectations before trying to write the first version.

I stopped giving every project or job equal space and used a few questions to decide what belonged:

  • What research problems do I want to work on?
  • Which past experiences prepared me to work on them?
  • What did I contribute or learn in those experiences?
  • Why is this particular program a sensible place for the next step?

I kept a common core but rewrote the ending for each program, connecting particular professors and labs to questions I had already introduced. This also helped me clarify which academic home I wanted. Starting early gave me something concrete, and initially bad, to improve.

The letters were the part of the application whose preparation went furthest back. When I contacted recommenders in early fall, I was relying on relationships built through research over months or years.

I wanted three professors who had supervised my work and could discuss how I thought, responded to setbacks, and contributed to research. Despite my industry experience at Meta, I did not use a manager’s letter. I wanted all three letters to address my academic research potential; that was the part of my work I wanted the committee to hear about from people other than me.

My part during application season was fairly practical: I asked people who knew my research, gave them time, and sent information about my work, programs, and deadlines. There were separate recommendation portals to track, too. The spreadsheet helped with that part, though the relationships behind the letters had very little to do with application season.

After submitting, there was little to do except wait. Programs followed different timelines, and I found it hard not to read something into every stretch of silence.

When I received an interview invitation from one lab I really wanted to join, I had so much nervous energy so I went for a run to clear my head. It ended up being 12 kilometers.

Most interviews felt like research conversations about my work, why I wanted a PhD, and which problems and professors interested me.

The process varied even within departments: one lab might arrange a conversation with the PI, another might begin with a student, and a third might schedule several meetings.

Preparing several lengths of explanation for my research helped me more than trying to anticipate every technical question. I read recent papers and prepared questions that could not be answered online, particularly about advising, collaboration, and what happened when research stalled.

Even with that preparation, I absolutely bombed a couple of interviews for programs I wanted. I wish I had been better at leaving those conversations alone afterward. It was easy to replay them and treat an uncomfortable interview as a verdict on whether I belonged in a PhD program.

In hindsight, I would have spent less of that waiting period replaying interviews and refreshing email, and more of it on research, friends, and exercise. The 12-kilometer run was probably a better use of energy.

After making offers, US programs often invite admitted students to visit. These are usually recruitment events: a chance to experience the culture, meet people, and imagine everyday life there.

I could not attend visit days in person because of the war with Iran. While other admitted students were walking around campuses and meeting potential labmates, I was trying to make the same five- or six-year decision from a bomb shelter. I was understandably preoccupied with the situation itself, but in hindsight I should have insisted on more remote interaction instead of accepting that I would simply have less information.

If I were making that decision remotely again, I would request more calls with potential advisors, their students, students who considered several of the same advisors, and people living there with partners or families.

I eventually chose between several programs with good research fit. Rankings did not separate them, so I thought more seriously about the life surrounding the PhD:

  • Where did I want to live for the next five or six years?
  • Where could my partner also be happy and build a fulfilling life?
  • What would our everyday life look like outside the university?
  • Which place offered the community and environment we wanted during such a significant period of our lives?

These questions were not secondary. My partner and I had to decide where we wanted to build a life; no ranking could decide that for us.

I would talk to students in the labs I was considering much earlier. Those conversations revealed more about advising and culture than any website.

I would also try to remember how little I could see of what shaped each outcome. I could research advisors, revise my materials, and think about where my partner and I wanted to live. Refreshing my email did not give me any more control over the rest.

I spent too much time worrying about admissions and not enough appreciating that I was already doing what I wanted to build a career around: research. Doing research had already shown me that I enjoyed the work and helped me build relationships with people who later supported me. Applications end; the research does not.

Some departments and student groups run pre-application mentoring programs. Their deadlines can be surprisingly early, so look well before applications are due.

Anonymous admissions forums are remarkably effective at turning a little incomplete information into a great deal of anxiety. People compare publications, interview dates, and GPAs while knowing almost nothing about one another’s research, letters, or fit. One stranger announcing an interview becomes “everyone has heard back except me.” If a site mainly encourages comparison with anonymous applicants, close it. I say this as someone who did not always follow that advice.

If you’re applying and have questions that aren’t answered here, feel free to reach out. While I certainly don’t have all the answers, I’m happy to share my experience if it might help someone else.