Beyond Quacking: Deep Integration of Language Models and RAG into DuckDB

2025/04/01 by Anas Dorbani, Dorbani, Anas, Sunny Yasser +5 · 9 voices · 5 citations
Computer Science · #Natural Language Processing Techniques #Topic Modeling #Speech Recognition and Synthesis

paper · pdf · doi:10.14778/3750601.3750685

Abstract

Knowledge-intensive analytical applications retrieve context from both structured tabular data and unstructured free text documents for effective decision-making. Large language models (LLMs) have significantly simplified the prototyping of such retrieval and reasoning data pipelines. However, implementing them efficiently remains challenging and demands significant effort. Developers must often orchestrate heterogeneous systems, manage data movement, and handle low-level concerns such as LLM context management. To address these challenges, we introduce FlockMTL: an extension for DBMSs that integrates LLM capabilities and enables retrieval-augmented generation (RAG) within SQL. FlockMTL provides LLM-powered scalar and aggregate functions, enabling chained predictions over tuples. It further provides data fusion functions to support hybrid search. Drawing inspiration from the relational model, FlockMTL incorporates: (i) seamless optimizations such as batching and meta-prompting; and (ii) resource independence through novel SQL DDL abstractions: PROMPT and MODEL, introduced as first-class schema objects alongside TABLE.

Citations

• Language Models are Few-Shot Learners

Cited by

• Rethinking Query Optimization for Multi-Agent Systems [Vision]
• SemBench: A Benchmark for Semantic Query Processing Engines
• Play by the Type Rules: Inferring Constraints for LLM Functions in Declarative Programs
• Research Challenges in Relational Database Management Systems for LLM Queries
• DBMS-LLM Integration Strategies in Industrial and Business Applications: Current Status and Future Challenges

Discussions

• Beyond Quacking: Deep Integration of Language Models and RAG into DuckDB arxiv.org/abs/2504.01157 github.com/dsg-polymtl/... [bsky, 22 points, 0 comments]
• Beyond Quacking: Deep Integration of Language Models and RAG into DuckDB https://arxiv.org/abs/2504.01157 (https://news.ycombinator.com/item?id=43616241) [bsky, 1 points, 0 comments]
• Beyond Quacking: Deep Integration of Language Models and RAG into DuckDB (arxiv.org) Main Link | Discussion [bsky, 1 points, 0 comments]
• Beyond Quacking: Deep Integration of Language Models and RAG into DuckDB #HackerNews https://arxiv.org/abs/2504.01157 [bsky, 0 points, 0 comments]
• Beyond Quacking: Deep Integration of Language Models and RAG into DuckDB https://arxiv.org/abs/2504.01157 https://news.ycombinator.com/item?id=43616241 [bsky, 0 points, 0 comments]
• Beyond Quacking: Deep Integration of Language Models and RAG into DuckDB https://arxiv.org/abs/2504.01157 [bsky, 0 points, 0 comments]
• ⚡ Hackernews Top story: Beyond Quacking: Deep Integration of Language Models and RAG into DuckDB [bsky, 0 points, 0 comments]
• Beyond Quacking: Deep Integration of Language Models and RAG into DuckDB https://arxiv.org/abs/2504.01157 (https://news.ycombinator.com/item?id=43616241) [bsky, 0 points, 0 comments]
• Beyond Quacking: Deep Integration of Language Models and RAG into DuckDB [bsky, 0 points, 0 comments]

Related

• Retrieval Augmented Generation (RAG) and Beyond: A Comprehensive Survey on How to Make your LLMs use External Data More Wisely
• Can Long-Context Language Models Subsume Retrieval, RAG, SQL, and More?
• When Retrieval Succeeds and Fails: Rethinking Retrieval-Augmented Generation for LLMs
• CoddLLM: Empowering Large Language Models for Data Analytics
• RB-SQL: A Retrieval-based LLM Framework for Text-to-SQL