

Shreya Shankar — Operationalizing Machine Learning
28 snips Mar 3, 2023
Chapters
Transcript
Episode notes
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Introduction
00:00 • 4min
Machine Learning and Practice Versus Studying Machine Learning in School?
04:03 • 2min
Is Database a Better Home for Model Training?
06:20 • 2min
How to Operationalize Machine Learning
08:16 • 4min
Is Distribution Shift a Problem?
12:14 • 3min
Is There a Difference Between Structured and Unstructured Data?
14:58 • 3min
Did Feature Stores Show Up Much in Your Interviews?
17:30 • 2min
Is Monitoring and Deployment a Big Pain Point?
19:04 • 2min
Monitoring for Data Corruption Is More Important Than Monitoring for Data Drift
20:45 • 2min
Is There a Context Behind This Corrupt Model Performance?
22:46 • 2min
Is There Still Data Corruption?
24:59 • 2min
Jupiter Notebooks - Is That a Good Idea?
27:14 • 3min
Is There a Trade-Off Between ML and ML?
29:55 • 2min
Jupyter Notebooks
31:43 • 2min
ML Tooling - What's Your Takeaway?
34:12 • 5min
I Think Reproducibility Is Incredibly Expensive to Get Perfect Reproducibility
39:00 • 2min
ML Engineers - What's the Difference Between Run Layer and Infrastructure Layer?
40:31 • 5min
MapReduce
45:18 • 3min
How to Hire a Data Scientist or ML Engineer?
48:35 • 3min
The Biggest Bottleneck in Machine Learning Is Interpretability
51:32 • 3min