Challenges in Building Production Systems with ML -- ML in Production Course @ CMU -- Lecture 1 скачать в хорошем качестве

Challenges in Building Production Systems with ML -- ML in Production Course @ CMU -- Lecture 1

This is the introduction lecture of the Machine Learning in Production course (17-645/11-695) at Carnegie Mellon University. Due to a recording problem, this is the Fall 2025 rather than the Spring 2026 recording. In this course, we focus on the engineering challenges when building applications and products with ML components, rather than just on how to build the models. This first lecture contains mostly a discussion of expected engineering and teamwork challenges in the context of a case study on Music Generation. We cut the syllabus discussion from this video. Course website: https://mlip-cmu.github.io/s2026/ 00:00 Welcome 00:47 Music Generation Case Study 06:03 Breakout Discussions: Challenges in Building a Music Generation Product 15:46 System Quality & Non-ML Components 19:53 Team Collaboration & T-Shaped People 27:28 What Makes Software with ML Challenging?