In-House Machine Learning Operations (MLOps) Platform
(Redirected from ML platform instance)
Jump to navigation
Jump to search
A In-House Machine Learning Operations (MLOps) Platform is an internal platform/ data processing platform system that supports the creation of ML systems (and their ML pipelines).
- Context:
- It can (typically) be used by an ML Engineer performing ML engineering, such as: ML model management, ML model training, ...
- It can be managed by an ML Platform Ops Team.
- It can be managed by an Internal ML Ops Team.
- It can range from being a Cloud-based In-House ML Platform to being an On-Prem In-House ML Platform.
- It can range from being an Custom ML Engineering Platform to being a Hybrid ML Engineering Platform to being a Pure 3rd-Party ML Platform, depending on how much it relies on a 3rd-party ML platform.
- It can contain ML Platform System Components, such as: an ML Workflow Platform, an ML Feature Store Platform, a Data Science Platform System.
- It can be related to a AI Platform.
- …
- Example(s):
- Custom In-House ML Platform, such as:
- Weights & Biases (W&B) Platform.
- ...
- a Hybrid ML Platform System, such as:
- Medable's ML Platform System, 2022 (based on GCP's VertexAI).
- SIE's Mastermind ML Platform System, 2022 (based on Databrick's ML platform).
- OpenGov's ML Platform System, 2016 a Spark-based ML platform (Spark ML).
- AT&T Wireless's ML Platform System, 2005 a SAS-based ML platform.
- …
- a Custom ML Platform System, such as:
- Viglink's ML Platform System, 2014 a custom ML platform.
- …
- …
- Counter-Example(s):
- See: ML Platform, ML Ops, ML Technology, MLaaS, Machine Learning Framework, Software Deployment, Software Development System, ML-based System, ML System Development.
References
2023
- https://jobs.netflix.com/jobs/278437235
- The Machine Learning Platform (MLP) provides the foundation for all of this innovation. It offers ML/AI practitioners across Netflix the means to achieve the highest possible impact with their work by making it easy to develop, deploy and improve their machine-learning models.
2018
- "Fact Store at Scale for Netflix Recommendations." Presentation at SparkSummit, 2018
- QUOTE: ... As a data driven company, we use Machine Learning algos and A/B tests to drive all of the content recommendations for our members. To improve the quality of our personalized recommendations, we try an idea offline using historical data. Ideas that improve our offline metrics are then pushed as A/B tests which are measured through statistically significant improvements in core metrics such as member engagement, satisfaction, and retention. The heart of such offline analyses are historical facts data that are used to generate features required by the machine learning model. For example, viewing history of a member, videos in mylist etc. ...