Biostatistics, quantitative safety science, global programme leadership through regulatory filing, and enterprise analytics. Then a doctorate to answer the question none of those roles could.
I joined pharmaceutical development as a biostatistician and worked my way toward the rooms where programmes get funded: quantitative safety science, then global programme leadership through Phase II, Phase III and regulatory submission, then enterprise analytics at Novartis.
The pattern repeated across companies. Governance spent enormous energy arguing how much each factor should count. Those arguments were serious, well prepared and genuinely contested. The decisions still came out much as they would have come out anyway.
The standard explanation is that the criterion was not weighted heavily enough. After enough cycles of watching that explanation fail to predict anything, I stopped accepting it.
Nearly every organisation frames it the same way: if a criterion matters enough, weight it heavily enough and the outcomes will follow. The framing is intuitive, it is almost universal, and it does not survive examination.
It turns out weighting is only one of three places a criterion can enter a decision, and often not the decisive one. By the time anyone is arguing about weights, the eligibility rules have frequently already settled which options could win.
That separation is the whole of my research, and the single most useful thing I bring into a governance discussion.
In 2022 I began a doctorate at ETH Zurich to find out whether the observation survived contact with evidence. For two years I ran it alongside a full-time role at Novartis. In 2024 I left the role to finish the work properly.
The three-way separation held. It also produced something I had not gone looking for: a result about self-set targets that requires no data at all, and that applies well beyond pharmaceuticals.
I defend in 2026.
The research →MBA, University of St. Gallen. MSc Medical Statistics. Twenty-three peer-reviewed publications and one book chapter. Based in Basel, Switzerland.
In a field where confident overstatement is normal, that is not modesty. It is the reason the verified claims are worth anything.
You will see it throughout this site: every number carries a source, and the limits of my own framework are on the research page rather than buried.
When a system proposes which options are worth considering, eligibility has moved. It was a rule someone wrote down and could point to. It becomes a process that is much harder to inspect. The argument about weights will carry on regardless. That is the part I find interesting, and the part I think is being under-examined.
I am not offering this as a service. I have not done the work yet, and saying otherwise would breach the only rule on this site I actually care about. When there is something to show, it will appear on the research page like everything else.
Ways to work together →