EPISODE 423 • NOVEMBER 19, 2026
Everyone wants to point AI at their lab data and get answers, but the data is rarely ready for it. Olga Vinogradova works on the informatics side of that problem, the work that happens before anything reaches a model. Jonathan Green talks with her about why uploading everything and letting AI sort it out does not work yet, how to rebuild data that nobody documented, and why the structure a lab chooses at the start shapes everything that comes after.
Key Takeaways:
Notable Quotes:
"Right now, half the things that I do is pushing a button and then watching it run, and then the other half is dealing with the alerts that come when you do that." — Olga Vinogradova
"Until it makes it to a SQL database, no amount of modeling is going to happen." — Olga Vinogradova
Connect with Olga Vinogradova:
LinkedIn: https://www.linkedin.com/in/ovinogradova
Enjoyed this? Follow The Artificial Intelligence Podcast and send it to the lab manager who still keeps results in a spreadsheet.