LLM for data analytics: text-to-sql 3 architecture patterns

preview_player
Показать описание
This is the first video in a series exploring how to work with structured data using Large Language Models (LLMs).

In this video, I explain the three main architectural patterns for building Text-to-SQL pipelines:

1. Prompt engineering & manual metadata retrieval (BASE)
2. BASE + RAG for metadata retrieval
3. 1 or 2 using the fine-tuned model

Stay tuned for more videos in this series on leveraging LLMs for structured data tasks!
  1. Denys on Data
  2. databricks
  3. llm
  4. llm sql
  5. rag sql
  6. rag to sql
Рекомендации по теме
LLM for data 0:06:59

LLM for data analytics: text-to-sql 3 architecture patterns

Natural Language Text 0:00:30

Natural Language Text to SQL: What are the Limitations?

How I Write 0:00:18

How I Write SQL using AI (Not ChatGPT)? Ai for SQL

New AI integration 0:13:19

New AI integration for your SQL databases | RAG, Vector Search, Admin Automation

Chat with SQL 0:58:54

Chat with SQL and Tabular Databases using LLM Agents (DON'T USE RAG!)

End To End 0:30:03

End To End Text To SQL LLM App Along With Querying SQL Database Using Google Gemini Pro

AI TOOL FOR 0:00:12

AI TOOL FOR SQL 🔥🔥 #sql #dataanalytics #ai

Text-To-SQL: LLM Generation 0:13:22

Text-To-SQL: LLM Generation With Labelbox

Building Knowledge Graphs 0:59:47

Building Knowledge Graphs w/ LLMs, NLP & Search w/ Alessandro Pireno - nyhackr June Meetup

A Natural Language 0:08:52

A Natural Language AI (LLM) SQL Database - Could this work?

Robust Text-to-SQL With 0:19:40

Robust Text-to-SQL With LangChain: Claude 3 vs GPT-4

Text to SQL 0:14:26

Text to SQL Agent for Data Visualization

Ask Your Database: 0:05:33

Ask Your Database: Build SQL AI Chatbot to talk to database

SQL AI Agents: 0:19:22

SQL AI Agents: Analyze Relational Databases with Natural Language using Llama 3 (LLM) and CrewAI

LLM-Powered Text-to-SQL with 0:18:30

LLM-Powered Text-to-SQL with Amazon Bedrock Agent Explained

AI tool for 0:00:19

AI tool for text to sql

Ep 12. Data 0:05:47

Ep 12. Data too Big for LLMs? Try Scratch Pads

Text to SQL 0:00:57

Text to SQL with NSQL from Numbers Station

How Does Rag 0:00:58

How Does Rag Work? - Vector Database and LLMs #datascience #naturallanguageprocessing #llm #gpt

Text to SQL: 1:31:30

Text to SQL: Building an LLM-Powered Analytics Dashboard for Your Company

ChatGPT In Excel 0:00:08

ChatGPT In Excel Shorts #2 #dataanalysis #excelfansonly #excelforbeginners #tutorial

Convert your text 0:00:21

Convert your text to sql with this AI tool #30dayschallenge #technology

LLM: Integrating Text 0:02:25

LLM: Integrating Text to SQL Using a Large Language Model

End To End 0:33:35

End To End Text To SQL LLM App | AI Project Tutorial | Text-To-SQL | Replace Data Analyst with AI

Комментарии
Автор

Great video. A few questions:

1) In the Final Prompt, we are including a lot of things. Question is how big a prompt can be?
2) In the base architecture, how is table/schema data fetched?
3) In the RAG method, are we doing semantic search to fetch only the relevant table/schema definition or do we fetch everything?

sarthak
Автор

A nice video, I just wondering in the part that checks if the query is validated. How I can do that?

Emiliano_Dev
Автор

Really good illustration Denys! Just one question, will this architecture still function well when you have too many tables with bad naming? I only see some products like AskYourDatabase work well with this situation. How should the solution fit in this architecture?

Jocob-Beller
Автор

Where can I find the flow diagram that you showed in the video?

utkarshshukla
welcome to shbcf.ru