There is an image that still persists in the collective imagination and that, let's be honest, many of us bought into when we started our careers: the romantic vision of the scientist. That solitary researcher, a sort of polymath genius who does absolutely everything.

In this narrative, the scientist is the one who designs the experiment, carefully cultivates their own cells in the incubator, secures the funding, analyzes every data point, and finally writes every word of the paper.

For decades, this model was the standard. But today, that glass ceiling has become evident. That romantic vision, while attractive, no longer scales.

The Trap of Technical Omnipotence

In the world of bioimage analysis, we live on the front lines of this reality. Today, it's not enough to be an expert in cell biology; if you want to do cutting-edge science, multidisciplinarity is not optional. You need to understand optics, statistics, and increasingly, programming.

The problem is that human knowledge has limits. When you try to cover all fronts—from maintaining cell cultures to optimizing deep learning algorithms—you become the bottleneck of your own research. Your project's progress is limited to what a single mind can process. And at the level of complexity we're working at in 2026, that's simply insufficient.

When Techniques Outgrow Individuals

Some bioimaging techniques have become so complex and multidisciplinary that they now demand multiple specialists to solve them. Spatial transcriptomics, for example, combines sequencing with microscopy, spot detection, and large-scale sampling. Each layer adds its own technical challenges, and no single person can realistically master them all without slowing the science down.

My Experience at the Karolinska: From "Doing Everything" to "Connecting Everything"

I remember when I arrived at the Karolinska Institutet, my mindset was still somewhat anchored in the idea that I should solve every technical problem that crossed my path. However, I encountered a reality that changed my professional compass.

I understood that my greatest value as a Bioimage Analyst was not trying to be the best in each isolated discipline, but rather acting as a connector of experts.

In a high-performance environment, success doesn't depend on how much you know individually, but on how well you can make the specialized knowledge of different people converge. My role shifted from "covering all fronts" to becoming the bridge between the biological question and the computational solution.

The Role of the Research Software Engineer (RSE)

Within this transition, a fundamental figure has emerged: the Research Software Engineer (RSE).

These are people who code with the rigor of engineering but with the heart of a scientist. An RSE doesn't just look for their code to "run"; they seek code that is:

  • Reproducible: Any other researcher can obtain your same results.
  • Robust: The software doesn't collapse at the slightest change in data.
  • Scalable: What started as a prototype on an afternoon in the lab becomes a solid tool for the scientific community.

A Necessary Cultural Shift

Fortunately, the culture is changing. Research groups are beginning to understand that hiring these technical profiles isn't a "luxury," but the only way to guarantee that science is solid and advanced. We're transitioning from the era of the scientist who does everything to the era of teams that connect everything.

At Phorma, this is our fundamental commitment. We believe that the future of science doesn't depend on overloaded scientists, but on the intelligent integration of many kinds of experts—software, data, microscopists, technicians, and more—within the flow of discovery.

Want to Dive Deeper into This Vision?

This point of view on scientists as connectors of experts is what sparked us to create Phorma. If you want to see how we translate that vision into practice, you can visit phorma.sh.

Sasha also wrote a another reflection on why we believe this is the path forward for science: https://ak.saxa.xyz/2026/02/18/por-que-cree-phorma-en.html