Almost everything I have built started with someone saying this takes too long.

Every role I have had turned into the same role. Two years at the United Nations Command in the Joint Security Area, where the entire job was accurate translation between four countries in real time, and where I rebuilt the workflows behind the operating log because they were losing information nobody could afford to lose.

At Bain, fifteen years of cohort economics rebuilt into three cases a client could actually decide on. At Relate, eighteen planned screens cut to the three features reps named, and a three-week build compressed into five days. At Capital One, fifteen leaders who each counted a head differently, and a six-hour forecast that now takes minutes.

Calibr is the most literal version of it. Tailoring a resume to one company took about two hours by hand and now takes twenty minutes, under a hard constraint: nothing in the output is invented, because speed that costs you the interview is not speed.

AI is why the last two moved as fast as they did, and I use it as a tool rather than a topic: acceptance criteria per screen instead of a description, a written rubric and a model grading a hundred runs instead of my own taste.

I finish at UNC Kenan-Flagler in May 2027 and I am looking for an APM role, ideally somewhere the product is still being figured out.

Four things I do, each linked to where I did it.

I write down what would have to be true.

Most bad product decisions are not bad reasoning. They are correct reasoning from a belief nobody wrote down, so nobody noticed it had never been tested. Listing the assumptions is cheap and it is the step that gets skipped.

I build the measurement before I trust the output.

Judging quality by reading it is a mood, not a method. On Calibr I wrote the rubric and had a model grade a hundred runs against it, which found a failure my own eye had been sliding past for weeks.

I think the smaller roadmap is usually the braver one.

Cutting scope reads as caution and is normally the opposite. Shipping seven things partly is the safe choice, because nobody can point at the thing you got wrong. At Relate that meant eighteen screens down to three features.

I bring the disagreement early, with the evidence attached.

An intern with an opinion is noise. An intern with a competitor teardown, user research and a smaller roadmap is making an argument the people who own the company can check. That is the Relate case in one sentence.