healthcare optimist | simply zero accomplishments

New York, USA
Joined April 2020
we've discovered teleportation and we're using it to get to the bus station on time. healthcare especially is full of assumptions we've stopped questioning. it's about time we picked some apart. a billion years since i last wrote something, so might as well have an upbeat title.
1
1
84
there’s some ideation on this in the healthcare proposals of Project2025
1
316
Replying to @anothercohen
think there are much better people out there for your “communion strategy”
78
combining local real-time de-id w/ re-id is pretty powerful you can utilize a 3rd-party app to improve your own service (currently, commonly a closed-source LLM) or covered entity end-users can use your service w/ clinical context below: diagram of potential workflow unlocks
2
1
100
going one step further - @NablaTech released an interesting post regarding their approach to privacy and security. other orgs seem to be doing this (@athelas) the data flow below articulates how one might re-identify to restore clinical context, post de-id
1
1
18
luckily, everybody's favorite hype cycle, the LLM, is great at de-identification in real-time some models are small enough to be installed locally
1
15
Cash Pay Transactions HIPAA strictly applies to specific electronic transactions. this means that strictly cash pay / direct primary care practices don't need BAAs. this is a pretty low TAM, and as soon as any electronic transaction occurs you're back at square 1
1
1
50
product-led growth is hard to come by in healthcare, with the divine hammer of privacy regulation waiting to smite you at almost every step. i thought through the question: where is it possible, if at all, to do bottoms-up distribution?
2
2
1
8
755
new post tomorrow! the topic: where in healthcare is it possible to do product-led distribution? i melted my eyes reading up on HIPAA, talked to policy + data experts, and put together my thoughts this was me reviewing to quadruple-check that I didn't say "HIPPA"
1
1
116