Why I Build Tools
software research
Every project I’ve worked on has hit the same wall: the analysis I need doesn’t exist yet. The data is too large, too noisy, or too weird for off-the-shelf solutions.
At first this felt like a detour — time spent coding instead of doing “real” science. But I’ve come to see it differently. The tool is the science. Deciding what to measure, how to normalize, where to draw the boundary — those are research decisions disguised as engineering ones.
Building your own pipeline also forces you to understand the problem at a level that no GUI ever will. You can’t write code to quantify something you haven’t precisely defined.
The best part? Once the tool exists, it doesn’t just answer your question — it lets anyone else answer theirs.
