Gabriel
Mongefranco

Architecture / Data / AI / Leadership

Connecting data, systems & people.Advancing Healthcare & Research.

I’m Gabriel Mongefranco, an enterprise and data architect with twenty years in healthcare, research, and enterprise technology. I turn the questions investigators, clinicians, and executives ask into data architecture, software, and teams that can answer them. Recent work includes a common data model for wearable research data at the University of Michigan, Fitbit sleep automation used by more than 15 studies, and readmission analytics that helped hospitals avoid more than $5 million in penalties.

Explore my work View my résumé
Staircase diagram of the standard mobile study data flow at the University of Michigan. Wearables and surveys feed a mobile app, which exchanges data with study platforms. Data then enters the U-M study database and flows on to analysis and visualization and to long-term storage. The database also feeds a data repository linked to the electronic health record.
The standard data flow for mobile and wearable research data at the University of Michigan, which I designed.
Experience across

Selected work

Systems with a purpose.

Interactive map showing which continuous glucose monitors, blood glucose meters, and insulin pumps work together, drawn as device cards joined by lines.University of Michigan / ResearchMobile & wearable dataA common data model and pipelines for research using connected devices.The university’s standard data flow for mobile research data, more than 100 technical articles, and open-source tools. Sleep automation used by 15+ studies and described in a 2025 paper in SLEEP.University of Michigan · 2023 – PresentRead overview In usePrivatium app launcher showing cards for Hello, Prescription Tracker, Animals, Sketch, and Pantry.Personal / Open sourcePrivatiumPersonal apps. Local data. Your hardware.Pre-release, version 0.3.2. Your records are plain text files you can read and back up, and the SQLite index is only a cache the program can rebuild.Rust · Lua · SQLite · 2026 – PresentView project In developmentDiagram of Extractium reading from websites, GitHub, library repositories, YouTube captions, local folders, and knowledge bundles, and writing a search index, llms.txt, SQLite, and Markdown outputs.University of Michigan / Research softwareExtractium™Builds a compendium, a portable static knowledge index any language model can use.One build writes a search index, llms.txt files, a SQLite database, and a Markdown folder. No server required.University of Michigan · 2026 – PresentRead overview In use

Shared foundations for research

So research teams do not start from scratch

The problem. Research teams were collecting wearable data study by study, through vendor APIs or exported flat files. That does not scale to hundreds of participants, it leaves large sensor datasets in files that are slow to analyze, and it makes NIH long-term preservation requirements hard to meet.

My part. I designed the standard data flow for mobile and wearable research data at the University of Michigan, the diagram above, and the common data model behind it, from the device and the vendor platform to a study database, analysis, long-term storage, and a repository linked to the health record. Then I wrote it down. The Depression Center’s Health Research Resource Library holds more than 100 technical articles I wrote on collecting and processing wearable data, and the center’s GitHub holds the open-source tools that go with them.

The result. A research team can start from a documented standard and working code instead of a blank page. The library and the code receive more than 500,000 views a year and have reached people in 188 countries.

From the notebook

Browse blog posts, poetry & more

Sometimes the simpler tool is the best answer to a complex problem.

In a university, the fastest route to adoption is a tool that needs no installation, no server, and as little new approval as possible, so I design toward that first.

  • Sleep data cleaning

    Problem
    A research lab cleaned Fitbit sleep data by hand, night by night.
    Usual answer
    A Python or R pipeline the lab would have to install and maintain.
    What I did
    Built the cleaning in Excel and Power Query, which most research teams already have, with a one-click refresh.
    Result
    Used by more than 15 research studies at multiple universities.

  • Collecting sleep diaries

    Problem
    Participants needed to report when they tried to sleep and when they woke.
    Usual answer
    A study app or a web survey with a login.
    What I did
    Participants text a shared mailbox, and a Power Automate flow parses the messages into a table the pipeline reads.
    Result
    No app and no account. The 2025 SLEEP letter describes timestamped texts as a low-burden method.

  • AI tools for sensitive research data

    Problem
    Researchers wanted AI assistance on files that cannot leave the institution.
    Usual answer
    A hosted AI service and a security review for every dataset.
    What I did
    Field Station AI runs language models in the browser from one HTML file, and Extractium™ writes its knowledge index as static files.
    Result
    No server, no database, no API. Institutional review still applies to regulated data, and the page says so.

How I build

Building with AI coding agents

Much of my recent code is written with AI coding agents, and I build the tooling that keeps that work contained and reviewable. Agent Airlock holds agents and their tools in a rootless container that sees only the project folders, not the rest of the computer. PeerFoil, still in development, adds independent review by agents from a different AI model family to agent-written changes.