At Postie, we set out to change direct mail’s reputation, and the team responsible for making it happen doesn’t get nearly enough of the spotlight: our data science team.
While the rest of the marketing world was busy building dashboards for search, social, and email, our data scientists were doing something harder: teaching direct mail to think. They’re the ones turning hundreds of millions of rows of customer data and thousands of behavioral features into the kind of precision targeting you’d expect from a digital ad platform, not a mailbox. Deep learning, random forest models, reinforcement learning, generative adversarial networks — this isn’t marketing jargon for our team, it’s the daily toolkit they use to build lookalike models, sub-segment audiences into meaningful personas, and predict which customers are ready to convert, when, and on what offer.
What makes this team special isn’t just the technical horsepower, though there’s plenty of that. It’s the combination of deep research chops and genuine obsession with the client’s outcome. Every model, every custom cluster, every AI-generated audience portrait exists to answer one question: how do we help a brand reach the right person with the right message, and prove it worked? Our data scientists don’t just build the models, they build the measurement infrastructure (matchback attribution, incrementality testing, real-time KPI tracking) that lets our clients see exactly what their spend is doing.
With nearly a decade of custom model development behind them, this team has taken data science work that used to take a full week of manual analysis and compressed it down to hours, automatically. They’ve shipped production ML systems that real clients depend on every day, and they’ve done it while staying remarkably grounded: the team’s guiding testing philosophy, an “exploit vs. explore” framework championed by our Head of Data Science, is as much about intellectual humility as it is about statistical rigor — always balancing what you know works against what you haven’t tried yet.
They come from wildly different backgrounds (physics, bioinformatics, pure curiosity) and they’d all tell you the same thing: the fundamentals matter more than the framework of the month, and the best idea is usually the simplest one you can explain clearly. That mix of technical depth and clear thinking is exactly why direct mail is finally getting the same intelligence and accountability that digital channels have had for years.
So without further ado, meet the people making it happen.
Dr. Nicholas Tyris, PhD – Head of Data Science and SVP of Strategy
Dr. Nicholas Tyris came to Postie from physics, chasing what he sees as the space race of the 21st century: AI and machine learning. The website’s platypus-in-space pulled him in, but he stayed for the chance to tackle genuinely hard technical problems and to put cutting-edge tooling against real business challenges.
As Head of Data Science, Dr. Tyris leads the team responsible for turning Postie’s direct mail platform into a genuinely data-driven channel — the applied research, custom modeling, and measurement infrastructure that sits behind every campaign. He’s also one of Postie’s most visible voices on the science behind the product, regularly presenting to clients and prospects on how machine learning applies to marketing.
Dr. Bryan Scott, PhD – Data Scientist
Dr. Bryan Scott is a Data Scientist at Postie and an astro-statistician whose career has spanned some of the most data-intensive research environments in modern science. He holds a PhD in Physics and Astronomy from the University of California, Riverside, where his research focused on intensity-mapping applications in galaxy evolution and modifications of Einstein’s theory of General Relativity. He was named the 2024 Hunstead Lecturer at the University of Sydney — an annual distinction reserved for internationally recognized scientists — and has held active memberships in three major cosmological research collaborations: CASTOR, the LSST Dark Energy Science Collaboration, and the Subaru Telescope Prime Focus Spectrograph Galaxy Evolution Survey Working Group.
Before joining Postie, Dr. Scott served as the Data Science Fellowship Program Postdoctoral Scholar at Northwestern University’s Center for Interdisciplinary Exploration and Research in Astrophysics (CIERA), where he also led data science training and education for the next generation of researchers. He holds graduate certifications in higher education pedagogy from UC Riverside and Northwestern’s Searle Center for Advancing Teaching and Learning, and previously trained and supported over 600 Teaching Assistants across UC Riverside’s academic programs.
Dr. Anthony Hodges, PhD – Data Scientist
Dr. Anthony Hodges is a Data Scientist at Postie and an experimental high-energy nuclear physicist by trade whose work has centered on extracting meaningful, rare signals from some of the largest data collections science has to offer. He holds a PhD in Physics from Georgia State University and has conducted research with major national and international collaborations including the PHENIX and sPHENIX experiments based at Brookaven National Laboratory, as well as the ATLAS experiment at CERN. His scientific work focused on rare, high-energy particle formations known as jets and their usage in understanding the behavior of nuclear matter under extreme conditions.
Before joining Postie, Dr. Hodges was an MPS-Ascend Fellow with the National Science Foundation and postdoctoral researcher at the University of Illinois Urbana-Champaign, where he served as a co-technical lead for the sPHENIX calorimeter group and helped lead the collaboration’s first public physics results. Across PHENIX, sPHENIX, and ATLAS, he worked with detector systems and analysis pipelines operating at the scale of hundreds of petabytes of complex experimental data, developing expertise in calibration, statistical modeling, uncertainty quantification, and large-scale data analysis.”
Learn From Our Team
This team’s best thinking doesn’t stay locked inside internal dashboards and client strategy decks. They’ve spent their careers turning impossibly large, messy datasets into signal, and they’re just as invested in helping other marketers learn to do the same.
That’s why we built Marketing Science 101 — and if this post made you want more of how this team thinks, this is where you go. It’s the same frameworks, testing philosophy, and modeling concepts they use on real client campaigns every day, taught directly by the people who built them and translated for marketers who want to think like data scientists without needing a PhD to do it.
You don’t need a technical background. You don’t need to read another dense research paper. You just need to register! Sign up for Marketing Science 101
