Python + MPP Database = Large Scale AI/ML Projects in Production Faster
summary
Shows that you can take Python ML into an MPP analytical database and move from prototype to production in minutes not months, using Vertica and the open source Vertica Python/VerticaPy interface. Gives concrete wins: Philips uses IoT telemetry and years of device history to predict component failure and schedule maintenance so machines never go down; Uber keeps hot data in a database across 100+ petabytes for sub‑second to few‑second queries and route planning under seven minutes. Claims one SQL call deploys models; community edition is free up to 1 TB and 3 nodes.
AI-written summary · about
chapters (13)
- 0:00Start
- 7:54Production Machine Learning Needs
- 9:38Advantages of Python Broad Utility
- 10:15Python Uses & Challenges
- 11:10End-to-End Machine Learning Process
- 12:00Challenges of Machine Learning at Scale
- 12:56Sampling vs. Full Dataset
- 14:07Data Has Gravity
- 16:15Uber Data Architecture
- 20:59Advantages of MPP Analytical Database
- 23:02Automate Model Training and Validation
- 23:32Deploy and Manage Model Life Cycle
- 24:59What People Are Saying
related talks
40mIn-Database Machine Learning with Jupyter
37mAccelerating Python Analytics by In-Database Processing
41mMassively Parallel Processing with Procedural Python
29mBlazing-Fast Python in Your Database: Unlocking Data Science at Scale with Exasol
37mIn database Machine Learning with Python in SQL Server
46mShorten the Path to Production with In-Database Machine Learning
33mYour best Bet: Effortless MLOps with Python Models in
30mEmpowered by Python - A success story