Data Engineering Vs Machine Learning Pipelines - What Is The Difference

แชร์
ฝัง
  • เผยแพร่เมื่อ 10 ก.พ. 2025

ความคิดเห็น • 10

  • @firefoxmetzger9063
    @firefoxmetzger9063 ปีที่แล้ว +1

    It is messy to directly compare feature engineering (FE) to the transform step in ETL; they exist on different levels of abstraction. A "traditional" ETL pipeline looks more like ETTTTL in practice because data is piped between multiple tables before it ends up in something like a dashboard or ML model. In the non-ML use case, we design that last "T" by consulting analysts and stakeholders to understand what subset of the data they need at which cadence for reporting/tracking/etc. In the ML use case, we design the last "T" using feature engineering to match the data to the requirements of the algorithm / ML model we are using.

  • @The-Rahko
    @The-Rahko ปีที่แล้ว +1

    Is it possible to become both a backend developer and a data engineer at the same time

    • @SeattleDataGuy
      @SeattleDataGuy  ปีที่แล้ว

      Its possible, but I guess I wonder what your goals are

    • @pragatitomar90
      @pragatitomar90 ปีที่แล้ว

      ​@@SeattleDataGuyi am doing Masters in CS ... is SWE - DE - MLE a good path or there are other quick stepping stones role to MLE?

  • @tonghongchen4289
    @tonghongchen4289 ปีที่แล้ว +2

    I found this video a bit abstract. Does anyone know a good comparison between Airflow and Kubeflow (or TFX)? That may help provide concrete examples of Data Engineering vs ML pipelines

    • @SeattleDataGuy
      @SeattleDataGuy  ปีที่แล้ว +2

      Oh that's actually a good topic, Airflow vs Kubeflow!

  • @mahmudhasan3093
    @mahmudhasan3093 ปีที่แล้ว +1

    I can tell all of that definitions you’ve put up there in the video is generated by chatgpt 😅

    • @SeattleDataGuy
      @SeattleDataGuy  ปีที่แล้ว +1

      Hmm, I am actually not sure. I'll ask Sarah though!

  • @Nick-du9ss
    @Nick-du9ss ปีที่แล้ว

    What is your opinion on Godfather of ai statement after resigned Google