Work With Me

I am always glad to hear from curious, motivated people who want to use data and models to improve public health decision-making. This page describes how I work, what opportunities are available, and how to get in touch.

Who I'm Looking For

I work with people at many stages — undergraduates, masters and PhD students, medical students and residents, infectious disease fellows, postdoctoral researchers, and collaborators from clinical, public-health, and computational backgrounds. The common thread is an interest in turning messy, real-world data into decisions and insights that protect people and animals from infectious diseases.

You don't need to arrive with a finished research plan or a specific technical skill set. The traits that matter most are curiosity, willingness to wrestle with imperfect data, and a commitment to doing work that holds up under scrutiny.

Prior experience with basic statistics and a scripting language (R, Python, Julia, or similar) is helpful but not required. If you are coming in without that background, we will teach you what you need along the way — plenty of people have joined this kind of work from clinical, laboratory, or policy backgrounds and become strong analysts. The non-negotiable piece is a willingness to work with these tools rather than around them. The research we do is heavily computational: most days involve writing and reading code, running and debugging models, and thinking carefully about numerical results. If that sounds like something you want to learn to do, even if you've never done it before, you will fit here. If it sounds like something you would rather avoid, this is probably not the right group for you.

What We Work On

My group at Wake Forest University School of Medicine studies the detection, surveillance, and dynamics of new and emerging infectious diseases. Representative projects include:

Outbreak response and rapid modeling. Building risk assessments and short-turnaround analyses to inform public-health response to measles, H5N1, mpox, COVID-19, and other emerging threats.
Antimicrobial resistance dynamics. Characterizing the spread of resistance in Neisseria gonorrhoeae, C auris, ESKAPE pathogens and others, and evaluating the downstream effects of prophylaxis, treatment, and prevention strategies.
Spatial and spatiotemporal epidemiology. Integrating mobility, climate, and social-vulnerability data to map risk and guide intervention planning.
Surveillance design and delay distributions. Quantifying underreporting, ascertainment, and the role of speed in detection and control.

Methodologically, we lean on Bayesian hierarchical modeling, mechanistic and statistical models of transmission, and reproducible workflows built in R, Julia, Stan, and Python.

How We Work

These are the working commitments I try to live by and that I ask people who join our group to take seriously. They are aspirations — I do not always get them right — but they are the culture we want to build together.

Chase the questions that matter

Our time is finite and the problems in infectious disease are not. It is tempting to pick projects because the data are handy or the method is novel, but the more important filter is whether the answer would change what a clinician, a health department, or a funder actually does. I would rather we struggle honestly with a question that matters than polish a tidy answer to a question no one was asking. When you start a project, be prepared to articulate — in a few sentences and without jargon — why it is worth doing and who benefits if we get it right.

Treat surprise as a gift, not a setback

In modeling and surveillance work, things will break: a pipeline will produce nonsense, a hypothesis will not survive contact with the data, a prior that seemed reasonable will dominate the posterior. This is the work, not a detour from it. An unexpected result is one of the best things that can happen on a given day, because the world just told you something you did not know. When you hit a wall, I would much rather hear about it early and plainly than watch a project drift. Post-mortems — what went wrong, what we would do differently next time — are something we do out loud, and I expect to do them on my own mistakes too.

Read widely and read seriously

The fastest way to become the kind of scientist who can tell a good question from a bad one is to know what other people have already tried. Expect to read — new preprints, old classics, work from adjacent fields, and the papers your collaborators are citing to you. A rough guideline: a few papers a week, carefully, plus a larger stack skimmed. When you start a new topic, most of your early weeks should be spent in the literature, not at the keyboard. Reading is not a break from research; it is research.

Be rigorous and reproducible by default

Public-health decisions get made on the basis of the analyses we produce, and those analyses need to hold up. In practice this means: code lives in version control from day one, analyses are organized so that a reasonable collaborator could re-run them without us in the room (and more frequently you yourself can re-run them in a few weeks or months from now), intermediate decisions are written down somewhere durable, and figures in any paper can be regenerated from a single entry point. When a result surprises us, our first job is to try to break it e.g., "jerk the steering wheel", by checking the inputs, testing the model on cases where we know the answer, and asking what else it could be. If a result cannot survive a determined attempt to falsify it, it is not a result yet.

Talk to each other honestly and kindly

The research will only be as good as the feedback we give each other on it. That means saying when something is not working, when an argument has a gap, or when a figure does not support its caption — and hearing the same from others without taking it personally. Feedback is about the work, not the person, and it is best delivered face to face and early. If something I have done is not sitting right with you, tell me directly; I will do the same. Praise is part of this too: if a colleague's talk, paper, or code made your week better, say so.

Keep one foot outside your project

Depth matters, but narrow depth makes dull scientists. Go to seminars that are not obviously connected to what you are doing, start or join a reading group, take an afternoon to play with a technique that has nothing to do with your immediate work. A surprising fraction of the best ideas in this field have come from people who borrowed a tool from a field next door. If you need help finding these groups, let me know; I'm here to help you develop in all aspects.

Work hard — sustainably

Science rewards stamina far more than it rewards sprints. The people who do great work over a career are not the ones who pulled the longest all-nighters; they are the ones who built habits that let them think clearly, most days, for many years. Figure out what that looks like for you and protect it. Take your vacations and actually unplug; stay home when you are sick; do not measure your self-worth in hours logged. If you are not happy with how your work is going, that is a conversation to have, not a reason to grind harder in silence.

Be accountable to the people we serve

The data we work with were generated by patients, clinicians, infection control specialists, other providers, and public-health staff who trusted that it would be used well. Much of the funding comes from taxpayers. Those commitments deserve to be honored in how we show up: arriving to meetings prepared and on time, flagging delays early, speaking up when we think something should be done differently, and finishing what we started. If you see something off — in my work, in your own, in a collaborator's — naming it is part of the job.

Mentoring Style

My goal is for every person who works with me to leave with a portfolio of work they are proud of and a set of skills that transfer beyond any single project. In practice, that means:

Regular one-on-one time. Weekly meetings are the default; additional time is always available when you are stuck or excited about something.
Real problems, real stakes. Projects are connected to active public-health questions, collaborators, or grants — not make-work.
Learning in public. I encourage writing (papers, blog posts, preprints), speaking (lab meetings, conferences), and sharing code openly through GitHub.
Ownership. Trainees are first authors on their primary work whenever possible, and I try to match the scope of a project to where you are in your career.
Honest feedback. I will tell you when something is not working, and I expect the same in return.

I also recognize that research is one part of a full life. I try to set expectations that are sustainable and to support people as whole humans.

How to Get in Touch

The best way to reach me is email: michael.dewitt@wfusm.edu.

When you write, it helps me respond well if you include:

A sentence or two about your background and where you are in your training.
What draws you to this kind of work, and any topics or methods that excite you.
What you are looking for — a rotation, a thesis project, a postdoc, a collaboration, informal advice.
A CV or short biosketch if you have one handy.

You don't need to have all the answers in your first message. A short, direct note is better than a long one you never send.

These working commitments are inspired in part by other groups who have thought carefully about how to do good science together — in particular, the Cobey Lab handbook at the University of Chicago, which I recommend reading.