Mission: build the infrastructure to make production healthcare databases AI-ready

[tmclerran]

Hi, I’m Tim. I’m a physician and my mission is to care for patients based on their own deeply informative data, with wisdom derived from a partnership between human and machine intelligence, trained on data from billions of other humans.

Healthcare data is a mess. To make healthcare data AI-ready, researchers have to spend a lot of time and effort curating dedicated databases separate from the databases that are being used clinically. By the time all of that is done, they can produce some really nice publications about what AI can do, and that’s usually the end of it. The gulf that divides these curated databases from the messy databases in clinical use is just too wide to cross, so their algorithms usually stay neatly tucked away in the pages of academic journals and never make it back to the bedside.

This problem arose because healthcare databases were generally created ad hoc as healthcare systems trudged through the arduous journey of digitizing their data. It was hard enough to keep important medical systems running smoothly during this transition, so nobody was thinking “and while we’re at it, let’s organize all these data in just the right way for AI to work, just in case we ever get computers powerful enough to do something with these data.”

Now we have the algorithms and the compute needed to do wonders with medical data. The next major step is to build live electronic health record databases that are AI-ready. Enter graphs.

There are many, many areas of the world that have not yet digitized their healthcare data. When they do, I would like to give them an AI-ready option.

If you’re interested, let’s talk.

Cheers,
Tim

http://linkedin.com/in/tim-mclerran

[clem]

I'm currently working at a company that's somewhat Healthcare related, although not at the level you're working with.

I do have an undergraduate degree in Biology (heavy on the biochemistry and genetics), so I have some sense of the biomedical field. Also, I took a Coursera class in Pharmacological mechanisms, which held a lot of surprises.

Good luck in your project!

[tmclerran]

Hi Clem! Thanks for the response. We've started a working group to create a proof-of-concept. If you're interested, shoot me an email at tmclerran@gmail.com and we can set up a time to discuss it!

Thanks,
Tim