To make the safest, most precise parenteral nutrition the standard for every newborn, everywhere.
Built by the researchers behind TPN2.0
Joe Phongpreecha, PhD
Co-founder & CEOJoe is a co-first author of the Nature Medicine study behind TPN2.0. He came to healthcare AI from machine learning in chemical engineering, and his focus is the hard translation work: turning the medical community's knowledge into a model, and that model into software clinicians can trust at the bedside.
Nima Aghaeepour, PhD
Co-founderNima is a Stanford professor of anesthesiology who leads the Aghaeepour Lab, with over 120 publications on AI for maternal and child health. He is the senior author on the TPN2.0 study and an inventor on the underlying patent, leading the science that keeps the model rigorous and continually improving.
It began in the NICU, with a hard problem
Total parenteral nutrition keeps the most fragile newborns alive when they cannot yet be fed. Yet the way it is prescribed is still largely manual, subjective, and slow. Small differences in a formula can change how an infant grows, and whether they develop serious complications.
Why newborns? Nowhere in medicine does one intervention buy more time. Save a newborn, and you don't add years to a life. You add a whole life, seventy-plus years of it.
We met at Stanford working on exactly this problem. Early on we made a deliberate choice. Rather than build one more model that predicts which babies will get sick, we set out to guide the intervention itself, because a prediction alone rarely changes what happens at the bedside. Over a decade of real neonatal orders, we learned how expert clinicians actually prescribe and where that practice quietly varies, and turned it into TPN2.0, which recommends safe, standardized, compoundable formulas inside the electronic health record with the physician always in the loop.
We validated it at a second hospital without retraining, ran a blinded study where clinicians rated it above current best practice, and published the work in Nature Medicine. Takeoff41 exists to take that research out of the paper and into everyday care, at any NICU that wants it.
Two beliefs shaped everything we built.
Clinicians stay in charge
TPN2.0 supports the care team, it never replaces them. Every recommendation is reviewable, adjustable, and physician-in-the-loop by design. We also build the way we believe: our products are always co-developed with providers in the loop, so they serve the real needs of the people who use them.
Evidence comes before intuition
We hold ourselves to the standard of published, peer-reviewed science, and we validate before we ship. Every claim we make about TPN2.0 traces back to data, and when the evidence points somewhere unexpected, we follow it rather than our assumptions.
