ML Platform Teams, Features Stores, and Where MLOps Extends DevOps With Aurimas Griciunas
🎞️ · 14.02.2023 · 18:35:50 ··· Dienstag ⭐ 0 🎬 0
📺 neptune_ai
🎬 · 14.02.2023 · 18:35:50 ··· Dienstag
😎 · 03.07.2024 · 15:40:29 ··· MiTTwoch
On this special episode of MLOps Live we have Aurimas Griciūnas as our guest. Aurimas shares his thoughts on ML platform teams, feature stores, versioning in data pipelines, and where MLOps extends DevOps with Piotr Niedźwiedź (Founder and CEO of neptune.ai).
Every couple of months, Piotr dives into a one-on-one discussion with an expert practitioner doing machine learning in production. Tune in for a no-fluff, in-depth MLOps conversation.
00:00 Introduction to MLOps Live
06:20 What is the difference between a data platform team and an ML platform team?
10:17 When is the right time to consider the right team for model development?
15:30 What are the responsibilities of both the data platform team and the ML platform team?
17:55 How can the data platform and ML platform teams collaborate in model development?
21:40 Why are pipeline models in production harder to monitor than software running on production?
26:17 Where do you store ML data for data versioning?
37:59 How do MLOps engineers train or retrain a model from scratch?
43:44 What are the training metrics for optimizing MLOps models?
48:00 Where does the computation level occur before calculating the features in a feature store?
Learn more about Aurimas Griciūnas:
Aurimas's Linkedin: https://buff.ly/3YtC1eH
Piotr's Linkedin: https://buff.ly/3RXIagI
Episode Resources:
Swirl AI newsletter: https://buff.ly/40RnZFk
neptune.ai blog: https://buff.ly/3f7wTpU
Designing Machine Learning Systems by Chip Huyen: https://buff.ly/3HYDTFg
Fundamentals of Data Engineering by Joe Reis, Matt Housley: https://buff.ly/3Inr4ph
Follow us & stay updated:
► Linkedin: https://buff.ly/41gEjzd
► Twitter: https://buff.ly/3n2Duer
► MLOps Community: (#neptune-ai)
► Github: https://buff.ly/41leUo0
SELECT * FROM DAvidPlay WHERE `DAvid`="6bfyfZFlT7M"
***
· 14.02.2023 · 18:35:50 ··· Dienstag
U
U
L
L
T
* 1676396150
* 1676396150
X 0
Y 0
P
C 144617
B 8
V 37
* · 21.03.2023 · 00:00:00 ···
2 · 29.05.2023 · 17:26:15 ···
L · 01.01.1970 · 01:00:00 ···
C · 01.08.2023 · 19:27:12 ···
💘 🖱️ * · 01.01.1970 · 01:00:00 ···
* · 01.01.1970 · 01:00:00 ···
· 01.01.1970 · 01:00:00 ···
**##
*** · 01.01.1970 · 01:00:00 ··· ::
*2* · 01.01.1970 · 01:00:00 ··· ::
*L* · 01.01.1970 · 01:00:00 ··· ::
*C* · 01.01.1970 · 01:00:00 ··· ::
********
