Colophon

About these notes

A working notebook from one job search and the interviewing that followed it, written down mostly so I would stop losing the same conclusions twice.

Why this exists

I prepared for a job search over about two months, went through several loops, and then — having landed somewhere — spent the following year on interview panels. Sitting on the other side changed my view of the process substantially, and most of what changed was not in the advice I had read while preparing.

These pages are the reconciliation. They are opinionated where I have evidence and hedged where I do not, and they are deliberately short. There is a great deal more that could be written about any of these topics; almost none of it would change what someone should do on a Tuesday evening six weeks before an onsite.

What this is not

• Not a question bank. There are excellent ones and I am not going to assemble a worse one. The pattern inventory on the coding page is a map, not a syllabus.
• Not company-specific. The loop described here is the large-company shape. Plenty of good places do something entirely different, and a take-home followed by two conversations is an increasingly common and arguably better process.
• Not a defence of the format. Interviews measure interviewing. The correlation with the job is real but modest, and treating a rejection as a verdict on your competence is a mistake I made repeatedly and do not recommend.
• Not advice for new graduates. The emphasis here on scope, ownership and design suits people with a few years behind them. Early-career loops weight the coding rounds far more heavily.

About the site

Static HTML and one hand-written stylesheet. No JavaScript, no analytics, no cookies, no fonts loaded from anywhere else, nothing to sign up for and no newsletter. It is a few pages of text; it did not need a framework and it does not need to know who you are.

The type is whatever your system provides. The colours follow your system's light or dark preference. Everything is one request per page plus the stylesheet, which is why it loads instantly on a bad connection — a property I appreciate on trains and see no reason to give up.

Corrections

If something here is wrong, I would like to know. Interview processes change, my sample is small, and several claims on these pages are generalised from a couple of dozen panels at one company — which is enough to form an opinion and not enough to be confident. Where I have stated something more firmly than the evidence supports, that is an error rather than a style.

Last revised in the spring. Back to the index.