Professor of Data Science

A Professor of Data Science divides time across three areas that academic institutions call the faculty triad: teaching, research, and service. The weight of each depends heavily on the institution. At an R1 research university, research output — publications, grants, doctoral students placed — is the primary currency of career advancement. At a liberal arts college or regional university, teaching quality and student outcomes carry more weight. Understanding which environment you're entering shapes every aspect of the job.

On the teaching side, courses typically include machine learning, applied statistics, data engineering, and domain-specific analytics. At the graduate level, seminar courses focused on current literature and methods are common. Course design at this level isn't just selecting a textbook — it means choosing which Python libraries to standardize on, which cloud platforms to require, which datasets tell the story of a method's limitations as well as its strengths. Good data science instruction requires updating material every one to two years as the toolchain shifts.

Research obligations at a research university are substantial and continuous. Publishing in top venues — NeurIPS, ICML, JMLR, KDD, or domain-specific journals depending on the faculty member's area — requires months of work per paper. Grant funding from NSF, NIH, or industry partners is increasingly expected, not optional, because it funds graduate student stipends and research computing resources that departments can't provide centrally. A professor without external funding is at a disadvantage in tenure review at most R1 programs.

Graduate advising is one of the most time-intensive parts of the job and one of the least visible to outsiders. A PhD student working on a dissertation needs regular technical feedback, help navigating setbacks in their experiments, guidance on academic norms, and eventually support finding a position. Advisors who do this well build research groups that amplify their own output; advisors who don't create students who stall out or leave the field.

Service — committee work, peer review, conference organizing, department administration — is the part of the job that expands to fill available time and is rarely rewarded as strongly as research output. New faculty are generally advised to limit service commitments until after tenure.

Education:

• PhD in statistics, computer science, electrical engineering, information science, or a closely related quantitative field (required for tenure-track positions)
• Strong dissertation research record; top programs expect candidates to have two to four first-author publications before finishing the degree
• Postdoctoral research experience increasingly expected at R1 institutions, particularly in competitive subfields like deep learning and causal inference

Research credentials:

• Publication record in peer-reviewed venues — the specific journals and conferences matter and vary by subfield
• Demonstrated ability to generate a research agenda, not just contribute to an advisor's existing program
• Grant-writing experience or co-investigator credits on funded projects are a meaningful differentiator
• Citation metrics are imperfect but are reviewed; a single highly cited paper carries more weight than several low-impact publications

Technical proficiency:

• Statistical modeling: generalized linear models, Bayesian inference, causal inference frameworks
• Machine learning: supervised and unsupervised learning, deep learning architectures, model evaluation and validation
• Programming: Python (NumPy, pandas, scikit-learn, PyTorch or TensorFlow), R for statistical analysis
• Data infrastructure: SQL, cloud computing platforms (AWS, GCP, or Azure), distributed computing (Spark)
• Reproducible research practices: version control, containerization, open-source code publication

Teaching and mentorship:

• Evidence of effective teaching — student evaluations, course design samples, or teaching portfolio
• Experience supervising undergraduate or graduate research projects
• Curriculum development experience, particularly for programs building or revising data science degree tracks

Service and professional engagement:

• Conference reviewing experience (program committee service signals professional standing)
• Department or graduate program committee participation during graduate training
• Industry collaboration or consulting history valued at professional-program-focused institutions

Academic positions in data science are in strong demand relative to most faculty disciplines, but the labor market is more nuanced than headlines about the data science boom suggest.

University enrollment in data science, statistics, and analytics programs has grown sharply since 2018. Many institutions that didn't have a data science department five years ago have one now, or are building one. That expansion has generated real faculty hiring — not just for research universities, but for regional institutions, community colleges offering applied certificates, and professional master's programs that need practitioners who can teach. The breadth of this demand is an advantage for candidates who don't fit the narrow R1 research profile.

At the same time, the academic job market for research-focused data science positions remains competitive. Top programs receive a high volume of applications for each tenure-track opening, and they are increasingly able to hire from a global candidate pool. A PhD from a well-regarded program with a strong publication record is necessary but not sufficient.

The salary gap between academia and industry remains substantial in data science specifically. A newly minted PhD in machine learning can expect offers in the $150K–$200K+ range from major technology companies, compared to $85K–$110K for an assistant professor position. This gap affects who pursues academic careers and means that universities competing for talent with industry backgrounds must offer compelling research environments, startup leave policies, or consulting flexibility.

For faculty already in established positions, job security at tenured institutions remains excellent. Tenure, once earned, is rarely revoked outside of serious misconduct. The question of long-term demand for data science faculty is tied to broader enrollment trends — if undergraduate enrollment declines accelerate, teaching-focused positions will face more pressure than research-focused ones.

The AI tooling shift is creating a parallel opportunity: many universities are hiring faculty specifically to develop responsible AI curriculum, lead AI ethics programs, or direct newly funded AI institutes. Faculty who can credibly bridge technical depth and societal impact questions are disproportionately valuable right now, as institutions respond to external pressure to address AI's broader implications.

Dear Search Committee,

I am applying for the tenure-track Assistant Professor position in Data Science at [University]. I completed my PhD in Statistics at [University] in May and am currently finishing a postdoctoral appointment at [Lab/Institute], where my research focuses on distribution shift in clinical prediction models and methods for uncertainty quantification under covariate shift.

My dissertation work produced three first-author papers, two published in the Journal of Machine Learning Research and one under review at NeurIPS. The applied thread running through all three is the same problem your healthcare analytics concentration is built around: prediction models that perform well on training data and fail in deployment because the population has changed. I've built open-source tooling around this work that has about 1,400 GitHub stars, which has helped me build connections with practitioners who are running into the same problems in the field.

On the teaching side, I designed and taught a graduate seminar on causal inference for observational data during my postdoc, with twelve students from statistics, epidemiology, and public policy. I restructured the first half of the course after the first two weeks when it became clear the students had very different baseline assumptions about what a causal claim requires — that experience taught me something about teaching across disciplines that I wouldn't have learned inside a statistics department.

I'm interested specifically in [University] because of the joint appointment structure with the School of Public Health. My research is genuinely interdisciplinary, and I'd rather be in an environment that makes that easier structurally than one where it requires constant negotiation.

I've attached my CV, research statement, teaching portfolio, and three writing samples. Thank you for your consideration.

[Your Name]