SD Labs is now live at labs.smartdata.net. It’s where we share what Smart Data people build between client projects and on their own time: internal software we now run the company on, open-source projects, side projects someone couldn’t stop thinking about, and the talks our engineers give.
Some of it is internal Smart Data work and some of it belongs to the people who built it, but all of it comes from people in our ecosystem. There are 15 projects on it today, and more on the way.
What’s on it
The projects fall into four kinds, and you can filter by each on the work page.
Built here. Tools we built for ourselves and now use every day. Our timekeeping, invoicing and project costs run on apps we built ourselves, written mostly by AI coding agents with a team around them. Our recruiting paperwork runs on an automation stack where you drag a candidate card and the documents send, sign and file themselves.
Open source. Projects anyone can use. Smartfire is a self-hosted workspace where people and AI agents share rooms, threads, boards and huddles, and our own team chat is migrating to it. ETLPlus is a Python ETL package and command-line tool, MIT licensed and on PyPI.
Side projects. Things people build because they want to. Ace Trace turns a short phone video of a tennis stroke into a technique report a coach can use. MockingBoard is an NFL Draft site where fans run mock drafts and write scouting reports, and Smart Data’s own fantasy football league runs on its fantasy site this season.
Talks. What our engineers present at conferences and community events, from three questions to ask before you build AI on a dataset to schema-grounded SQL generation for enterprise analytics. Our medallion architecture session at Taste of IT on November 18 is on there too.
Why we’re sharing it
Most of what we do for clients stays private, and it should. That makes it hard to show how our people work, and Labs fills that gap. The engineers who build data platforms and AI systems for clients are the same people who build these tools, and the habits carry over: start from the real workflow, keep the data trustworthy, and own what you build.
It also says something about where software is heading. Several of these projects replaced software we used to pay for. With AI coding agents doing much of the typing, a small team can now build and run tools shaped exactly to how a company works. We’re using that on ourselves first, and it changes the build-or-buy question for a lot of teams. If that’s a conversation you want to have, our page on how we build with AI is a good place to start.
How it stays current
Projects get added as they ship, and the Lab log shows what changed and when, newest first. Each project page says what it is, how it was built and who it’s for, with a link to the code where the code is public.
You’ll also meet a small pixel character around the site. He shows up on project cards and in the launch video, and he’s the closest thing Labs has to a mascot.
Take a look
Start at labs.smartdata.net and click into anything that looks interesting. If you’re an engineer who likes building this kind of thing, we’re hiring. And if you see something you’d want built for your team, let’s talk.
Questions this raises
What is SD Labs?
SD Labs (labs.smartdata.net) is where Smart Data shares what its engineers build between client projects and on their own time: internal tools the company runs on, open-source projects, side projects and conference talks.
Is SD Labs a separate product or company?
No. It is part of Smart Data. Nothing on it is sold as a product; it shows the work of the same people who deliver data platform and AI projects for clients.
Can I use the open-source projects?
Yes. The open-source projects, such as Smartfire and ETLPlus, are public, and each project page links to its code and license.
Can Smart Data build something like this for my team?
Yes. Several Labs projects replaced software Smart Data used to pay for. If your team is paying for software it could own, talk with us about what that would take.
