From notebooks to production - bringing devops to scientists
summary
A deployable model-serving stack called Assaf turns a notebook function into a running service via a single-library import, uses RabbitMQ queues and connectors for request/response and a high-performance collector on Kubernetes, and ships built-in tracing and metrics plus CI/CD (GitHub + Drone + Docker). A deployed NLP model initially takes ~60s and after real use is optimized to under 1s, showing the system works and drives prioritized engineering effort.
AI-written summary · about
chapters (6)
- 0:00Intro
- 0:28What happens when an ID is born
- 4:55Darwinian production
- 10:10Continuous Integration
- 12:28Webbased Processor
- 15:09Soft Platform
related talks
28mDevOps for Data Science With Kubernetes
45mArchitecture of an NLP Deployment
33mDeploying and Scaling Spark ML and Tensorflow AI Models
44mFrom an AI Model in a Notebook to a Real-Time Service: Let’s Architect!
54mEconomics and Best Practices of Running AI/ML Workloads on Kubernetes
16mKeynote: Accelerating AI/ML Development with Open Source Software
35mCase Study: AI-as-a-Service on Kubernetes at Scale and In Production
01h 00mFrom notebook to production using