These are the pieces we intend to write first, drawn from work already in production. They are listed so the shape of the section is clear — none of them exists as a post yet, and none is dated.
Carriers do not fill in forms. What it takes to turn “we can do that one for 1,275 all in, pickup Thursday morning” into a structured bid a dispatcher can trust — and why the original message has to stay visible next to it.
A query that asks for 25,000 rows, silently returns 1,000, and looks complete. How we hit it, how we now catch it, and why “it looked fine” is the most expensive sentence in a data pipeline.
The brand, the campaign, the sample messages, the opt-in language, and the queue you do not control. A plain account of the compliance surface behind business text messaging in the US.
We have shipped systems we cannot put a number on, because nobody measured the before. What we do differently now, and what a useful baseline looks like when the process being replaced was never instrumented.
Why we put the repository, the database and the deploy pipeline in your name from the first commit, what it costs us, and what it means when an engagement ends.