The Future of Data is Words
summary
SQL celebrates 50 years while LLMs push natural-language queries; demos show Flink SQL stream pipelines that convert multi-currency order streams into unified orders, a Steampipe engine exposing AWS and Slack APIs as SQL tables, and GPT‑4o auto‑documenting a catalog. These work as proofs of concept but fail at enterprise scale without three pieces: retrieval/RAG to pick relevant datasets, a deterministic semantic layer and knowledge graph to map business metrics, and rich metadata to disambiguate terms like “active users.” Companies that build high‑quality metadata win.
AI-written summary · about
chapters (11)
- 0:00Welcome and opening
- 1:09SQL turns 50
- 3:05Magic demo with SQL
- 4:31History of data querying
- 8:52SQL standard adoption
- 9:56SQL in the internet era
- 11:48Modern SQL use cases
- 15:28AI era and natural language
- 19:53Future of semantic layers
- 26:42Predictions for the future
- 29:35Closing and Q&A
related talks
15mFrom Words to Data: Using LLMs to Turn Questions into SQL
41mData Engineers: Respect the Resiliency of SQL
12mPower a Better Future
15marchive – Manohar Sai Jasti
29mRevolutionizing Text Transformation.
31mFrom Data to Precision: Fine-Tuning LLMs for Proprietary Query Languages
40mOne SQL to Rule Them All
44mNonrelational SQL