Text2SQL: The Dream versus Reality - Laurel Orr | Stanford MLSys #89

Episode 89 of the Stanford MLSys Seminar Series!

Text2SQL: The Dream versus Reality
Speaker: Laurel Orr

Abstract:
In this talk we explore the nuanced application of Large Language Models (LLMs) in Text2SQL for data-centric tasks. The talk outlines the transformative potential of LLMs in making data insights accessible through natural language, while critically examining the hurdles such as the 80% solution dilemma, the translation of business terminology, and privacy concerns. Through the lens of the DuckDB-NSQL case study, we proposes solutions like a semantic layer and modular pipelines, showcasing a novel approach that leverages an intermediate language for enhanced SQL compilation. This session not only highlights the technological advancements but also the pragmatic challenges and strategies for deploying LLMs in real-world scenarios.

Bio:
Laurel Orr am currently a researcher at Numbers Station part of the Numbers Station Labs where she think about all things foundation models and data tasks.

Before Numbers Station, she was a PostDoc at Stanford working with Chris Ré in the Hazy Research Lab. In August of 2019, she graduated with a PhD from Paul G Allen School for Computer Science and Engineering at the University of Washington in Seattle, where she was part of the Database Group and advised by Dan Suciu and Magdalena Balazinska.

--

Stanford MLSys Seminar hosts: Avanika Narayan, Benjamin Spector, Michael Zhang

Twitter:

--

#machinelearning #ai #artificialintelligence #systems #mlsys #computerscience #stanford

[muhannadobeidat]

Good presentation, everyone that tried this reached similar conclusions. It is great to see that confirmation and similar though process here

[kenchang]

I enjoyed the discussion and experience sharing. Thank you very much.