Back

Speaker "Peter Relan" Details Back

 

Topic

Leveraging Google's 2018 release of BERT Deep Learning Engine to deliver on the promise of Natural Language to SQL.

Abstract

In the NL to SQL space we have developed the first commercial product with our deep learning model based on Google BERT's Transfer Learning technology for Natural Language understanding and trained it and tuned it for generating SQL queries on Cloud Data Warehouses.

Who is this presentation for?
Analysts, Data Scientists, Business Intelligence leadership

Prerequisite knowledge:
Understanding of what a database is.

What you'll learn?
How fast the field of self-service analytics is evolving with the help of AI and addressing the problem BI backlogs.

Profile

Peter Relan comes from the internet research community (working as a research engineer for one of the co-authors of TCP/IP after his undergraduate work at UCLA) and the database industry (working at Oracle with database and SQL gurus after his graduate work at Stanford). He has been involved in numerous start-ups including Business Signatures (SQL for Stream processing), ThreatMetrix (Fraud Detection), Discord (Slack for 300M gamers) and Got It. Peter's AI interests are in the Natural Language to Program Generation space, specifically using Deep Learning Neural Networks and generating SQL from English questions for self-service BI>