Being six months into starting a company aimed at providing scientific services, I want to take some time to explain to clients, collaborators, friends, and family what led me down this path.
The obvious alternative to the scientific start-up is pursuing an academic career - an option that contains quite great freedom to pursue knowledge. So why not chose the academic option?
As some of you know, a career in academia can be a pretty risky proposition - if you fail to advance to tenure you are often forced into short-term contracts and uncertain employment. But starting a company is not the way to get stability. Rather, the uncertainty increases even further. A well-defined research project from a passionate researcher about can mostly get the needed research grants - at least in Luxembourg. But a company can fail in two ways, fail to find funding or fail to be profitable in the long run.
No, the reason to create a startup is not for stability. It is to execute a mission that academia is ill-equipped to handle - problems that are easy in principle, but not in practice. If a problem is easy to solve in principle, no grant application proposing a solution will be successful.
Providing simulations for Scanning Tunneling Microscopy(STM) is one such easy-in-principle-hard in practice problem. In principle, the Tersoff-Hamann approximation from 1983 is easy to apply and sufficient in many contexts. In practice, inexperienced researchers often misapply the method due to common numerical pitfalls. Or they struggle to find the computational resources needed to perform the heavy calculations.
A private company, in contrast, thrives on standardized problems - on always pushing for higher efficiency. Doing so provides consistent benefits for all clients and allows the company to grow.
And to succeed, a company must spread its solution as far as possible.
All in all, the incentives of a private firm fit best with my mission to bring simulation services to all STM (and SPM) experimentalists. You can take a look at the first product in the pipeline and see what I have come up with.