Microservices for Machine Learning
summary
Describes building recommendation services that use machine learning on user actions to recommend ads, and the architecture challenges of distributed training pipelines. Covers using traditional microservices for deploying model results to production and less-traditional microservice patterns for organizing model training, and contrasts ML engineering approaches with conventional software development.
AI-written summary · about
chapters (5)
- 0:00Introduction
- 6:05Recommendation microservices
- 16:48Offline model pipelines
- 31:01Key lessons
- 33:05Q&A
related talks
41mRecommendations for Building Machine Learning Software
21mHarnessing Microservices for Scalable, Fault-Tolerant ML Systems
35mIntegrating a Spark ML Pipeline into a High-Load Low-Latency Recommendation Service
37mMLOps Architecture for a Recommendation Platform
48mPrinciples and Practices of Building Machine Learning Systems
39mDeveloping Machine Learning for Impact
26mPatterns of ML Models in Production
41mInfrastructure for ML Applications