Charge Your Database with Predictive ML Capabilities using Open-Source
summary
Claims that a database can run ML as virtual tables and return forecasts via SQL, implemented with MindsDB’s AI Tables that use the Lightwood engine and can run in Docker or MDB Cloud. Demonstrates SQL like create predictor to train models from a source table, batch predictions via joins, scheduled refresh jobs, and OpenAI/GPT connectors. Notes scarcity of experts (~1% of devs know ML), shows working predictions for rental prices and a practical, maintainable workflow that actually returns live forecasts.
AI-written summary · about
chapters (6)
- 0:00AI and ML challenges
- 3:10AI integration in software
- 10:58Why databases for ML
- 15:12Demo: AI tables in MindsDB
- 22:31Benefits of database ML
- 26:01Q&A
related talks
21mWhat If We Could Use Machine Learning Models as Tables?
22mPredict the future (or don't) with machine learning in applications
12mOpenMLDB: An Open-Source Real-Time Feature Platform
45mForget SQL! Create your own AI Data Analyst
40mIn-Database Machine Learning with Jupyter
55mDistributed In-Database Machine Leaning
31mIn-database machine learning: Automate, scale, deploy
33mYour best Bet: Effortless MLOps with Python Models in