Generative AI Building Blocks

This repository contains a collection of building blocks for generative AI use cases.

The first use case we are going to tackle is Text2SQL. We are going to build an application that can convert natural language questions into SQL queries using LlamaIndex and Claude 3 through AWS Bedrock.

1. Text2SQL using LlamaIndex and Bedrock using Candle Glow dataset Text2SQL using LlamaIndex and Bedrock using Star Wars dataset

2 - TBC

Environment setup

pyenv install 3.12.0
pyenv local 3.12.0
python -m venv .venv
source .venv/bin/activate

Start a local docker Postgres Instance

docker run --name movies-db -e POSTGRES_PASSWORD=mysecretpassword -d -p 5432:5432 -v $(pwd)/postgres:/var/lib/postgresql/postgres:rw postgres:alpine
docker run --name candle-glow-db -e POSTGRES_PASSWORD=mysecretpassword -d -p 5432:5432 -v $(pwd)/postgres:/var/lib/postgresql/postgres:rw postgres:alpine

Interactive bash shell

docker exec -it movies-db bash
docker exec -it candle-glow-db bash

Connect to Postgres

psql -h localhost -p 5432 -U postgres

Create a database

CREATE DATABASE star_wars;
CREATE DATABASE candle_glow;

Connect to the database

\c star_wars
\c candle_glow

Use your SQL client like DBeaver to connect to the database

Once connected, open a new SQL script and run the following SQL (./llama-index/text-to-sql/star_wars_db/database.sql or ./llama-index/text-to-sql/candle-glow/database.sql) to create the tables and insert some data.