Dave's PostgreSQL Stuff

Ollama is an open-source project that simplifies running large language models (LLMs). It runs locally on your machine, and you can choose from multiple LLMs. Keeping all the data on your machine should provide a security bonus.    DBeaver is a universal database tool with an AI Assistant. This assistant provides an extra layer of security by allowing you to lock down what is shared, with the default being metadata only, no data. Webinar I recently presented a webinar  on using the DBeaver AI Assistant. Several questions were raised about using Ollama, one of the many AI Engines available. But I had not tried it. The first step is downloading Ollama for your platform (Linux, Windows, or Mac) and installing.  Once you have Ollama installed, you will need to pick a LLM.  I chose the Gemma3 and installed it using the command line client. Adding the Gemma3 LLM to Ollama DBeaver Configuration Ollama listens on port 11434. I was able to load the Gemma3 model us...

 Structured Query Language, or SQL, is my favorite programming language, but its quirky syntax can sometimes be frustrating. Many have professed a need to bypass SQL and query databases in English (another quirky language. Is it possible to have an Artificial Intelligence take a request in English and translate it into SQL on my behalf? I am using the AI Chat Feature of DBeaver Enterprise to test.  The Prompt  My prompt was  Create a table in this postgresql database to hold address book information and then add ten example records from random data to OpenAI.  The Response Sure, let's create a table named `address_book` with the following fields: `id`, `first_name`, `last_name`, `email`, `phone_number`, and `address`. CREATE TABLE address_book (     id serial PRIMARY KEY ,     first_name varchar ( 50 ),     last_name varchar ( 50 ),     email varchar ( 100 ),     phone_number v...

 If you've ever wanted to skip SQL and its quirks and write queries in English, I am happy to report that you can do it with an AI of your choice and DBeaver. I had a webinar on July 24th that shows you how to do just that.   I will show you how to set up DBeaver  using a wide variety of commercial AIs to let you work with an AI. We will also go through the security features to ensure your data is safe. By default, only the schema metadata is shared with the AI. You can lock down access to just one or more tables being available to the AI. Imagine not having to translate a data request from management speak to SQL! Here we simply ask an AI to provide the five most popular films by revenue and get an answer.  I will also show you how an AI can help by explaining and fixing query problems. You can even identify how that 'thrown over a wall' query works. Have you ever been handed a strange database and have no idea of its content? Then, use an AI to identify its u...