Investigative Data Journalist
Independent investigative reporting on corporate accountability in U.S. healthcare and regulated industries — built from primary documents and the federal datasets most newsrooms never open.
Primary sources, always. Every fact starts in a document I have read myself.
Investigations
Ongoing, document-driven investigations into how power organizes itself to avoid being measured. Described here at the level of the beat — the specifics stay in the reporting until they are published.
How private-equity owners route money and liability out of the homes that care for the elderly — a facility's ownership followed through the shell entities in CMS federal ownership data, related-party rent verified at the worksheet level in HCRIS Medicare cost reports, the federal wrongful-death docket read line by line.
A longitudinal read of ICE's own detention-management data, documenting the quiet migration from a phone check-in app to the GPS ankle monitor — and following the money to the contractors who profit from the switch.
OSHA fatality and catastrophe inspections across four states, set against 29 CFR 1904 — the rule that leaves the employer to decide whether a death on a data-center site ever becomes a federal record at all.
A new safety signal surfacing in the FDA's own FAERS data, cross-read against a drug's label-revision history and its maker's SEC disclosures — the distance between what the data shows and what investors are told.
In progress. Tips welcome — see Contact.
Method
I came to journalism from data and software. I now point that toolkit at institutions that would rather not be measured — which means I do not wait for a leak. I go and get the records.
Every fact begins in a primary source — a court filing, a cost report, an inspection record — read directly, never through a press release or a summary.
I pull and analyze federal datasets few reporters open: Medicare cost reports at the worksheet level, payroll-based staffing files, ownership disclosures, contract records. When the data does not come clean, I write the code to make it usable.
I use automation and AI to read at a scale a person cannot — whole dockets, thousands of filings — to find the anomaly worth chasing. Every published claim is then verified by hand.
Nothing runs as fact unless a primary document supports it. Contested claims are attributed to their source and carry the response of the party they concern.
Every named party gets a genuine, specific opportunity to respond before anything is published — and their answer is reported alongside the finding.
About
Powerful institutions are required to write things down. Cost reports, court dockets, inspection files, adverse-event databases — the record of what an institution did, and often of what it would rather you never notice, is already public. It stays hidden only because reading it is slow, technical work that almost no one does. That work is the whole of my beat.
I report on corporate accountability in U.S. healthcare and other regulated industries, and I build every story from the underlying documents — never a press release, never a summary. Federal court dockets pulled through PACER and RECAP. Medicare cost reports read at the worksheet level in CMS's HCRIS system. Payroll-based staffing files, ownership and enrollment records, OSHA fatality inspections, FDA adverse-event data, EPA and state environmental filings. The primary record, line by line.
The investigations that hold me share one shape: real harm, deliberately made hard to see. How private-equity owners route cash and liability out of the nursing homes that care for the elderly — legible in the shell entities of federal ownership data and the related-party rent Medicare disallows on Worksheet A-8-1. How immigration enforcement migrated from a phone app to a GPS ankle monitor, and what the contractors running it promise their investors. Who is cited, and who never is, when a worker dies on a data-center construction site. A drug-safety signal surfacing in the FDA's own data quarters before anyone writes it down.
The method underneath is plain and non-negotiable. Nothing runs as fact without a primary document behind it. Contested claims are attributed and carry the answer of the party they name. Every institution I write about gets a real chance to respond before I publish. The rigor is not ornament — it is what separates an investigation from an accusation.
The toolkit comes from an earlier life in data and software, which is an advantage rather than a detour: I can pull a federal dataset most reporters never open, and write the code to make it usable when it arrives broken. I studied rhetoric and writing at the University of Texas at Austin, and I report from the Texas–Mexico border — a dual citizen, working both sides of the line in either language.
Contact
For editors, sources, and collaborators. I read everything myself and respond quickly.
For editors
Available for commissions and pitches on healthcare, private equity, and regulatory accountability.
For sources
If you have documents or knowledge of wrongdoing, I protect the people who come forward. Reach me first by encrypted email; for anything sensitive, ask and we will move to Signal. I will never name you without your consent.
Based on the
Texas–Mexico border · Reporting in English and Spanish