SQL That Doesn't Suck & A WebDev Future with Claude

Last month was an amazing turn out of 35 people. This month we have two talks involving databases.

SQL That Doesn’t Suck (Your Bandwidth or Your Soul) - Bill Sunderland
We all know how to write `SELECT * FROM myTable`, but what happens when your app starts dragging like it’s stuck in molasses? This talk is about writing SQL that’s lean, mean, and ready to serve — even when business requirements change faster than your coffee cools.

We’ll explore how to pull just enough data to keep your frontend happy (whether it’s Python, Node, or something else), without overwhelming your database or your users. You’ll learn how to turn vague business requests into precise queries, avoid common performance traps, and keep your app snappy and scalable. Bonus: no actual souls will be harmed in the making of these queries.

Bio:
Hi, I’m Bill — a proudly neurodivergent technologist who thinks in pictures, builds in SQL, and organizes like a spreadsheet with feelings. I’ve architected data systems across industries, but healthcare is where I’ve found my groove. My brain works like a puzzle: I see the big picture, spot the missing pieces, and build solutions that actually fit. I communicate in bullet points, diagrams, and the occasional Star Wars reference. I ask a lot of questions — not because I don’t get it, but because I want to get it right. If it touches data, I’ve probably worked with it. If it involves community, I’m probably building it. And if there’s a Super Smash Bros. tournament, I’m probably winning it.

A WebDev Future with Claude
We will start with an unknown system that has TSQL as part of its business logic. Claude and I will interact to:

• Comment the code - this will allow me to understand the code and ensure Claude does too.
• Format the code with a linter - this will allow me to read it better as a human (and the 200+ other files I have to do this with)
• Ask Claude to propose a test plan.
• Write tsqlt (SQLServer unit test framework) tests for the file and properly infer where they should go in my project.
• Run tsqlt to show the file and tests are properly created
• We'll take a short detour into Flyway and talk about versioned vs. repeatable migrations in the context of developing stored procedures.
• Work our way through a couple of the tests to ensure the logic is sound.
• Use and MCP Server to ask Claude to tell us where the data elements are being used in the system and how.
• Put up a draft PR in Github.
• Pretend to edit the PR accordingly.

Claude will write nearly all of the code, but I will gain most of the understanding of this legacy system our client needs to modernize. I maintain my responsibility.

Hi, I’m Bill — a proudly neurodivergent technologist who thinks in pictures, builds in SQL, and organizes like a spreadsheet with feelings. I’ve architected data systems across industries, but healthcare is where I’ve found my groove. My brain works like a puzzle: I see the big picture, spot the missing pieces, and build solutions that actually fit.I communicate in bullet points, diagrams, and the occasional Star Wars reference. I ask a lot of questions — not because I don’t get it, but because I want to get it right. If it touches data, I’ve probably worked with it. If it involves community, I’m probably building it. And if there’s a Super Smash Bros. tournament, I’m probably winning it.

Bio:
It's Boyd. Meh.

Agenda
5:45pm - 6:00pm - set up and socialize a bit
6:00pm - 6:45pm - SQL That Doesn’t Suck - Bill
6:50pm - 7:45pm - A WebDev Future with Claude - Boyd
8:00pm - #BeerOps