Dr. Thomas Kelly
speaker
542 appearances
1 recordings
1 series
first heard Jul 2026
last heard 27 Jul
Dr. Thomas Kelly’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsRecordings per month over the last 12 months — 1 in all, peaking in Jul 2026 with 1.
Appearances
Um, now obviously we have to be careful.
Like we're not trying to replace the clinician, but it has to be in a way that's useful for the clinicians.
Um, so whether that's, I don't know, like a place where patients get a summary of their visit and then they can ask some questions and we can send them to the next doctor that uses Heidi or something.
I can imagine us doing that.
Um, and then one other interesting area, which is a bit of a tangent, is um hardware.
So we're actually spending quite a bit of time because we own the models.
This is back to medical.
device.
So we fine-tune open source models um to create all of Heidi's uh notes.
That's important because back to the X-ray analogy, we need to know what's in the box.
Like it can't be like Anthropics API because we don't know what's behind that.
So we don't we don't believe that regulators in the end will trust like a third party API.
Like I can't say my you know my medical device is uh GPT 5.5.
And they say, well, how can you guarantee that it performs?
And we say, well, OpenAI says it's going to do the same thing.
And you know, we know from our data that it doesn't.
Um, so yeah, once you have the models, then you can actually run it on premises.
So you can build hardware where they can, they don't even need to share their health data with Heidi.
They can just run it locally in their hospital.
Um, this is a bit more RD, but we're excited about the potential of that because there's certain juristics.
Showing 301–320 of 542 · page 16 of 28
← Previous
Next →