How to be fast and have a good memory with Machine Learning
summary
MemSQL presents a main-memory database with built-in TensorFlow integration, Kafka-connected create pipeline primitives that attach streams to tables, and transaction-safe exactly-once semantics for upserts. The engine is production-ready (core shipped by 2013; nearly $100M raised) and runs both on-prem and in a private cloud service; a Thorn deployment compresses investigations from months to two days and helps find around 2,000 children in 2016.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 1:05Machine Learning
- 1:50Machine Learning and Spark
- 3:03Realtime Data Pipeline
- 4:47Live Duplication
- 5:27Onpremise vs Cloud
related talks
05mTeaching databases to learn in the world of AI
05mDriving the On-Demand Economy with Predictive Analytics
06mImplementing custom models for machine learning
12mOpenMLDB: An Open-Source Real-Time Feature Platform
23mAccelerate the Machine Learning Pipeline on Very Large Datasets
43mNatively Engineered for AI: The Evolution
17mMachine Learning at the Speed of Light with Spark
26mSpeed up Your ML Workloads With Kubernetes Powered In-memory Data...