↓Skip to main content

MLaaS vs MLOps

·1 min

I watched this interesting video regarding the differences of MLOps (e.g. Train, Serve, Monitor your own model) vs ML-as-a-Service (e.g. Foundation Models).

MLOps vs ML-As-Service // Jill Chase and Manmeet Gujral

These are some interesting learning points:

  • MLaaS tend to be more B2C and MLOps is mainly B2B
  • Foundation Models have pushed the MLOps Frameworks to emphasize on the software’s ability to perform fine-tuning
  • MLOps revenue generation is not as fast as MLaaS but the contracts and projects are bigger
  • MLOps tends to lean towards providing for instances of Proprietary Data.