"You are your own best postdoc."

When I talked to established PIs, one piece of advice came up again and again:

"You are your own best postdoc."

Looking back at this year, I think I finally understand what that means.

During my postdoctoral years, I spent a lot of time thinking about projects. Not just experiments for the next paper, but larger questions. Questions that I found interesting but could not easily address at the time because the infrastructure, methods, or resources were not available.

Like many postdocs, I accumulated a mental list of ideas. Some realistic, some ambitious, some perhaps slightly crazy.

Most of them stayed exactly that: ideas.

Starting my own group changed this.

For the first time, I could build workflows specifically around the questions I wanted to ask. We invested a lot of time into setting up screening pipelines, automating routine tasks, establishing data analysis workflows, and creating the infrastructure needed to generate large and reproducible datasets.

The interesting thing is that "being your own best postdoc" does not necessarily mean doing the most work yourself.

For me, it means that I can finally test some of the ideas that have been following me for years.

At the same time, I feel a strong responsibility towards the people in my group. When students or group members invest months of effort into a project, I want the work to have a high chance of producing robust and interpretable results. Because of that, many of the projects we do in the group are built around questions where I feel confident that the approach is sound and the outcome will be informative regardless of the exact result.

That leaves me with a slightly different role.

I often end up taking on the higher-risk ideas myself.

These are projects where I genuinely do not know whether the approach will work, whether the hypothesis is correct, or even whether the question can be answered in the way I imagine. They are the projects that kept me thinking during my postdoc years. The ones for which I previously lacked the infrastructure, throughput, or technical setup.

Some of them will fail.

That is part of the process.

But when they work, they can open entirely new directions.

This year, much of my time was spent building exactly those capabilities. The automation side came relatively naturally. I had already spent years thinking about workflows and experimental design, and I genuinely enjoy organising large-scale experiments. There is something deeply satisfying about seeing hundreds of plates move through a screen and having everything come together at the end.

The bioinformatics side was more challenging. I had enough knowledge to understand what I wanted to achieve, but not enough to get there efficiently. The first scripts took far longer than they should have, often involving hours of debugging for seemingly simple problems. But over time, those workflows became reliable, and the analyses that once took days now run almost automatically.

What I learned from all of this is that, at least for me, being in the lab is still an important part of being a PI.

Everyone runs their group differently, and there is no single right approach. But at this stage, I enjoy building new things, testing ideas directly, and working on questions that I have been carrying with me for years.

And perhaps that is what "being your own best postdoc" means to me.

Not that I do everything myself.

But that, after years of wondering "what if?", I finally get the chance to find out.