Tales from the Field: The Future is Now
summary
Emerging work shows machine learning’s payoff depends less on models and more on data logistics — about 90% of effort goes into preparing, versioning and serving data and models — so teams adopt data ops and DevOps patterns, keep many specialized tools, and treat big data as simply data. Real systems use Apache Flink for true streaming, Kafka or MapR for persistent stream transport, Apache Drill for schema-on-read exploration, and Kubernetes for orchestration, enabling a stream-first architecture and hybrid cloud deployment.
AI-written summary · about
chapters (13)
- 0:00Intro
- 0:42The Future
- 2:20The Future is Here
- 5:08Big Data
- 9:48Processing Data
- 15:58Flexibility
- 16:58Streaming
- 19:53Traditional Retail
- 22:08Stream Replication
- 23:53No sequel databases
- 25:58Data orchestration
- 28:13Machine learning and AI
- 31:08Machine learning in AI
related talks
11mRecent trends in data and machine learning technologies
47mExpert Panel: Current & Future State of the Art in Real-Time Data Infrastructure
16mLife is But a (Data) Stream: Building Quality Data Pipelines
31mThe Future of Data on Kubernetes: From Database Management To AI Foundation
57mBACK TO THE FUTURE: EMERGING TRENDS IN DATA ENGINEERING
32mBig Data Processing : Where do we go next ?
41mInfrastructure for ML Applications
34mTales of Data Architecture Evolution