• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
  • HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI

    Shreya Shankar on AI Functions Entering SQL Mainstream

    Researcher notes AI operators in SQL now operational after earlier skepticism toward DocETL.

    SS
    1 Source, 27d ago, first seen 27d ago

    TLDR

    Shreya Shankar, a databases and HCI researcher building DocETL, stated that AI functions are going mainstream. She recalled that several AI colleagues had dismissed the project two years earlier, claiming better models would handle unstructured data processing on their own. Shankar observed that AI-powered operators are instead being added directly to SQL and expressed satisfaction at the shift. The post also notes her upcoming role as assistant professor at CMU CS in Fall 2027.

    Combined views

    15.8K

    1 Source, first seen 27d ago

    Combined views

    15.8K

    1 Source, first seen 27d ago

    105 likes
    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    105 likes
    14 comments
    102 saves
    11 reposts
    14 comments
    102 saves
    11 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    1 Source

    @sh_reyaLove that AI functions are going mainstream. Several AI folks laughed at us when we started building DocETL; saying that better models are going to “solve” unstructured data processing. But 2 years later joke is on them: AI-powered data processing is being operationalized, as AI-powered operators (AI functions) in SQL are making database vendors tons of money. I think we are only at the tip of the iceberg and there is *so* much interesting stuff to be done all across the stack, from BI/interfaces for humans and agents to figure out the right questions to ask, down to custom LLM inference/execution engines and new features and indexes for storage engines

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    1 Source

    @sh_reyaLove that AI functions are going mainstream. Several AI folks laughed at us when we started building DocETL; saying that better models are going to “solve” unstructured data processing. But 2 years later joke is on them: AI-powered data processing is being operationalized, as AI-powered operators (AI functions) in SQL are making database vendors tons of money. I think we are only at the tip of the iceberg and there is *so* much interesting stuff to be done all across the stack, from BI/interfaces for humans and agents to figure out the right questions to ask, down to custom LLM inference/execution engines and new features and indexes for storage engines