Truveta wins AI innovation award at SXSW

Last week, Truveta was recognized with the South by Southwest (SXSW) Innovation Award in AI. I love that our extended team of Truvetans and health system members were recognized for their years of work which have gone into building Truveta so far—and there is so much more to do on our mission of Savings Lives with Data.??

The potential of medical AI is written about endlessly these days – but AI is only as good as the data it is trained on. I honestly don’t know of any medical dataset like Truveta Data which is representative of the diversity of our country, timely, complete, and clean. We look forward to partnering with all these medical AI solutions to help them deliver regulatory-grade solutions.?

Using AI to unlock the most complete, timely, and clean EHR data???

The award specifically went to the Truveta Language Model, which we use to “clean” the unstructured medical data. The only thing structured in healthcare today is your medical bill, which says nothing about health outcomes. It says nothing about side effects and all those notes your doctor is typing when you visit them about where the pain is, why you stopped taking a particular medication, or your family history.??

I’m often asked why this hasn’t been done before, and I see two challenges we’ve needed to overcome to make this possible.??

First, the data is fragmented over all these providers. We all go to multiple providers. It took a pandemic to get a set of providers to say enough is enough and come together, pooling our data so that we can create a dataset that is representative of the whole country so that we can start finding the insights to take care of our patients.?

But second, it also now requires this incredible AI technology to process this massive volume of unstructured data. The average hospital produces ~50 petabytes of data each year. These are doctor’s notes and images. This data is fragmented, inaccessible, and unstructured. ?

The pandemic was a moment in time that inspired unprecedented action by leaders of several innovative health systems who came together to create a data set we can all learn from. And now you have this technology – cloud computing and AI – that make it possible.?

Lifting the hood on TLM?

At SXSW, it was fun to really geek out with some of the surprisingly savvy AI attendees. Some of my favorite discussions walked through our data engineering console to show the distributed data flow, or the AI ops console which tracks quality of the model over time vs continuously growing evaluation datasets (discussed in our TLM whitepaper ).??

These are the good old days!?

I often reflect on the Bill Gates quote – “We overestimate what will happen in the next 2 or 3 years. We underestimate the next 10.” Now that we have all this data prepared by AI, ready for training new AIs, what will we accomplish over the next 10 years???

P.S.??

We had some fun in Austin, too. ???