Explore the challenges and solutions in scaling MLOps for computer vision in this 59-minute conference talk from the MLOps Community Mini Summit #4. Learn about Flyte, an open-source platform for agile AI product development, and its application at Recogni for automotive perception stack development. Gain insights into running AI models at scale from Blackshark.ai's experience in analyzing global satellite imagery. Discover how to manage large datasets, ensure model reliability, and maintain system performance in production environments. Hear from industry experts David Espejo, Fabio Grätz, and Arno Hollosi as they share their experiences and lessons learned in implementing MLOps practices for computer vision applications.
Scaling MLOps for Computer Vision - MLOps Mini Summit Meetup 4
Overview
Syllabus
[] Introduction to David Espejo, Fabio Grätz, and Arno Hollosi
[] Flyte: A Platform for the Agile Development of AI Products - David Espejo
[] Flyte at Recogni - Fabio Grätz
[] Lessons Learned from Running AI Models at Scale - Arno Hollosi
[] Other examples of computer vision-heavy startups in the automotive world uh that are using Flyte and production
[] Attraction of Flyte
[] Wrap up
Taught by
MLOps.community
