Jekyll2026-09-01T08:14:04+00:00https://danieledegregorio.dev/feed.xmlDaniele De GregorioSystems Engineer at KLA specializing in electron-beam metrology at nanometer scale. Physicist with expertise in nanophotonics, machine learning for inverse design, and experimental particle physics.Daniele De GregorioA lab notebook, on the web2026-08-19T00:00:00+00:002026-08-19T00:00:00+00:00https://danieledegregorio.dev/blog/a-lab-notebookThis site started as a single-page portfolio. That is still the front door. The blog is a place to write things down that do not belong on a résumé card: how a measurement was actually made, what broke, and what I would do differently next time.

I work on electron-beam metrology systems. The interesting part is rarely the headline number. It is the chain underneath — source, column, stage, detector, model, and the statistics that decide whether a result is real.

What you will find here

  • Notes from building and calibrating precision hardware
  • Short write-ups on nanophotonics and inverse design when I have something concrete
  • Occasional systems-engineering asides: interfaces, error budgets, and why “it works on my bench” is not a spec

I am not aiming for a publication venue. If a note is useful to one other person who has stared at a noisy residual and wondered whether the tool or the sample is lying, that is enough.

Precision is a social contract with the next engineer who has to trust your number.

A typical first check when something drifts is embarrassingly small:

residual_nm = measured_nm - model_nm
print(residual_nm.std())

If the scatter is honest, you can talk about the mean. If it is not, you go back to the instrument.

More soon.

]]>Daniele De Gregorio