* Pipelines for research are different from those end-to-end ones in production.
* We are all doing ML R&D the wrong way. The point is to recognize the practices that will survive for the next decade.

I never did get around to find out if anyone ever reads these things. Tell me if you did? Thanks.
" /> * Pipelines for research are different from those end-to-end ones in production.
* We are all doing ML R&D the wrong way. The point is to recognize the practices that will survive for the next decade.

I never did get around to find out if anyone ever reads these things. Tell me if you did? Thanks.
" />

Improver

Contact Details

E-mails & Phone

Ariel Biller

Ariel Biller

Evangelist | @LSTMeow | Renowned MLOps Shitposter

Tel Aviv

Dad, Researcher-turned-Evangelist, deconstructor with a heart for optimization.
Argues well enough on the following:
* MLOps is a mission statement, not a fancy name to "DevOps for ML".
* Pipelines for research are different from those end-to-end ones in production.
* We are all doing ML R&D the wrong way. The point is to recognize the practices that will survive for the next decade.

I never did get around to find out if anyone ever reads these things. Tell me if you did? Thanks.

Experience:

  • Senior meta-Realist (Hour One)
  • Data Science Lead (StoreDot)
  • Evangelist (AllegroAI)
  • Deep Learning Research Scientist (AllegroAI)
  • MLOps Coordinator (BriefCam)

Skills:

People also viewed these profiles

People also viewed these profiles

Dan Engel

Regional Sales Director at Poly @ Ramat Gan, Tel Aviv

People also viewed these profiles

Roman Belinsky

Scrum Master at Dalet Digital Media Systems @ Southern