About · The house
Voxel · since 2024
Built by people who read.
Voxel wasn't designed in a meeting room. It was written between shifts, by radiology residents who felt every day the distance between the rigor a report demands and the tools that existed to produce one.
2024 · The front line
While the world argued over whether artificial intelligence would replace the physician, we were on the front line. Residents at HCFMUSP, reading shift after shift, seeing radiology's real pain up close: too many tools, descriptors scattered everywhere, laterality errors, and precious time lost on what matters least.
The build
There was no tool worthy of the mind behind it. We put together a technology team and started building what we missed: a way to read that doesn't hold you back, that opens possibilities instead of imposing constraints. Every product decision came out of something that bothered us in the reading room — not out of a spreadsheet.
The reincarnations
We rethought the platform many times. What began as a library of findings became, after countless reincarnations, a reporting hub where the radiologist reads however they see fit. It emerged at USP, tested in the residents' everyday work, and today it reaches the market to offer a new way of producing reports, with more freedom and more efficiency.
“We made it for ourselves first. Then we understood it was for you too.”
The team
Founders · InRad · HCFMUSP
Guilherme Nakandakare
MD, University of São Paulo · Radiology resident at InRad · HCFMUSP
Iago de Paiva D'Amorim
MD, Federal University of Juiz de Fora · Radiology resident at InRad · HCFMUSP · Former coordinator of the InRad AI Group (2025–2026)
Silvério Shindi Hosomi
MD, University of São Paulo · Radiology resident at InRad · HCFMUSP
Rian Radeck
Computer Science, UNICAMP · MSc in AI, Institut Polytechnique de Paris · CTO at Cogni
What drives us
Four principles · no exceptions
I · Principle
The radiologist at the center
The tool serves the mind operating it — never the reverse. Authorship, style, and clinical judgment are yours; the platform follows.
II · Principle
Report-grade rigor
We hold every screen to the precision a good report demands. Consistency isn't the user's effort; it's the tool's doing.
III · Principle
Built in the reading room
We build between shifts, testing on real days. What doesn't survive a real day of reading doesn't make it into the product.
IV · Principle
Quiet to think in
An interface that gets out of the way. Fewer alerts, less friction, more room for the reasoning only the physician can do.
Born in the DNA of the University of São Paulo.
The USP DNA Seal is granted by the USP Innovation Agency only to companies proven to have been born inside the university. It can't be bought. It has to be proven.
USP Innovation Agency · InRad · HCFMUSP
The story continues in your next read.