The record
- Now
- Senior Data Engineering Consultant and AI Capabilities Lead
- Before
- Options Market Making Firm, Chicago, 2015–2020
- School
- University of Michigan, BSE Industrial and Operations Engineering
- Based
- Metro Detroit
- Age
- Thirty-three
The work
Six years at a consulting firm. I joined as employee number five; it is around twenty people now. Before that, five years at an options market making firm in Chicago.
Four of my engagements have been for private-equity-backed healthcare companies: senior housing, hospice, optometry and dermatology.
The role is worth stating precisely, because people in my position routinely overstate it. I did not source those deals, value those companies, or decide what to buy. The sponsor decided. I came in afterward and built the reporting and the data infrastructure that let them actually run what they had bought.
How I work
Nobody working inside a job has the room to stop and ask why it still takes an hour. There is always something more urgent that week. Asking it anyway is most of what I do.
The answer is usually not dramatic. Somebody learns one new thing, or takes the extra few days to build it properly once instead of doing it by hand forever. The hard part was never the technology.
So I am not arriving to transform anything. The improvements are small, there are a lot of them, and they point in the same direction. Stacked up over a year, that is what changes how a company runs. Taken one at a time, none of them is frightening.
Whether I got it right is measured by what people do with the time afterwards. I want them spending it on the work they care about and are good at — the reason they took the job, and the reason they are still there years later.
How you go about it matters as much as what you build. You treat people with dignity and respect and put yourself in their shoes. You also hold a standard and do not let it slip. Most people are good at one of those. Running a company somebody else built means both, on the same day.
Some of what that has looked like
I turned a week of hunting into a job that runs itself.
Four accountants had been reconciling the general ledger against the data warehouse line by line, every month, so the company could stand behind its financial filings. I built the thing that isolates every variance to the line item that caused it. It took a month to build.
Thirty to fifty hours a month back to the revenue accounting team — and they stopped staying late four nights a week.
I closed the month for forty-six properties reporting fourteen different ways.
Fourteen operators, each sending their numbers in their own format, all of it landing on one fund that needed a single set of books. It had been about fifteen hours of manual work every month.
Every property now reconciles to zero variance before anybody looks at it.
Two companies merged and their billing systems did not speak.
One firm bought another, and each logged time its own way. Every hour worked had to reach the right invoice without somebody typing it in twice. I built the middleware that kept the two in step.
Nobody rekeyed an hour of work while the two companies became one.
I built the scorecard that thirty weekly meetings ran on.
Four hundred metrics, generated automatically, across a firm of two hundred and twenty-five people. It was the piece that moved the place onto running by numbers.
Every meeting started from the same set of figures, and nobody had to build them first.
I taught myself the tool so I could lead the rebuild.
They had outgrown their reporting platform. The right rebuild needed something I had not used before, so I learned it and then led the build rather than hand it to somebody else.
They came off the old platform onto a warehouse they could extend themselves.
Eight advertising platforms, one set of numbers.
Spend was scattered across eight ad and analytics platforms that each counted things differently. I own the architecture that brings them together, and I wrote the checks that verify every pull independently rather than trusting the pipeline.
The reporting flags its own failures before anybody acts on a wrong number.
If you’re years away from selling, that’s the normal case. Book it anyway.