At Redgate, we’re experimenting with how AI can help developers and DBAs become better at their jobs. Everyone is asking for AI, as well as the ability to turn AI off. We’re working hard to accommodate both requests as we look for ways to leverage AI.
One of the areas we’ve started to add some AI is in Flyway Desktop (FWD), with a few features designed to help reduce the cognitive load and save time as they work with databases.
I’ve been working with Flyway and Flyway Desktop for work more and more as we transition from older SSMS plugins to the standalone tool. This series looks at some tips I’ve gotten along the way.
Summarizing My Changes
Commit messages are hard. There are numerous posts on the art of writing one, as well as no shortage of examples showing poorly structured messages. As I look through my own messages in the SQL Saturday repo, I’m somewhat surprised at times at ow well I’ve described something, or how poorly. I’ve certainly had the “fix” messages creep into my work.
When we start making database changes, we often have a number of things we might touch. A new column could necessitate view and stored proc changes. One view might impact another, or a function. A piece of work might include alterations of a few tables, and fixes that might not be obvious need a bit of a description in the commit message.
If nothing else, we need to ensure we’re linking commits to planned work on Azure DevOps boards, in Jira, or wherever a project manager might be assigning tasks.
Flyway has added an AI feature to help us summarize our changes. It’s in preview as of Jan 12, 2026, and is designed to help ensure we get a better commit message than “fixed.”
Generating Commit Messages
When I go to the VCS tab in Flyway Desktop (FWD), I see all my changes listed. In this case, I’ve been demoing and testing a lot of things that I haven’t committed. Bad practice, and I’m a little upset with myself for not keeping my repo clean.
Notice the little sparkle below the Commit button? This is where the AI help lives. If I select a few files, in a few seconds I’ll get a message generated for me. This is a decent summary, at just over 50 characters.
If I don’t like this and want to edit it I can. If I do that, the AI won’t overwrite this. I’ll change the wording and select another file. Notice that the message remains the same.
If I want a new message, I can click the AI sparkle and it will regenerate this. I’ll show this in the video below.
If I select all the files, the AI summary isn’t great, but then again, I don’t know how I’d summarize all these changes. I might not do any better.
This is a small thing, but I often find that I struggle to come up with a meaningful message at times. Sometimes I do a good job and add something relevant with a work item number. Sometimes I want to just type “new table” or “fixed bug”.
This reduces some of the creativity burden when I’m working through database changes.
Enabling AI Features in Flyway
This is a preview feature as of Jan 5, 2026 as I write this. To get this in your FWD, your organization needs to have enabled AI features in your portal. I’m just a member, but whoever is an admin for your Redgate products would find it here.
In FWD, you need to look at the Preview Features item under the config menu.
In this area, you can enable or disable features as needed. I see these marked as Red-gate only, but I think they are supposed to be released to some customers by this time. It’s likely I need to upgrade my FWD, which I’ll do when I have time.
Once you do this, you should start seeing some AI stuff with the purple/pink shaded area and the sparkle icon that we’re all seeing everywhere.
Summary
Getting meaningful commit messages can be hard, but it’s also a tedious thing. I find that developers often get lazy and do the minimum. That AI might not be much better, but it does reduce the need for a user to think of a message and gives them a starting point from which to work.
To me, this is one of the better uses of AI, handling a common, but tedious task. Now to go clean up my repo and take advantage of this with a bunch of smaller commits.
This feature is documented, but we are likely to enhance and change it a bit, so all feedback is welcome. If your organization doesn’t want you using AI, and you could share some schema from a migration script, I’d be happy to test it for you and see what summary is produced and send it back to you. Ping me on X/LinkedIn/BlueSky if you want.
Flyway is an incredible way of deploying changes from one database to another, and now includes both migration-based and state-based deployments. You get the flexibility you need to control database changes in your environment. If you’ve never used it, give it a try today. It works for SQL Server, Oracle, PostgreSQL and nearly 50 other platforms.
I wrote about Side Jobs a few weeks ago and today I’m taking vacation from Redgate for mine. I’m at Colorado Crossroads, a large volleyball tournament in downtown Denver. I suspect my players are more excited than I am since they get to play and miss school.
Hopefully there’s plenty of coffee as I flew back from the UK yesterday.
Redgate recently released SQL Compare v16, which included a new feature to work with Redgate Data Modeler. I decided to give it a try in this post. I’ll take a model and compare it to a database, and deploy my model.
There’s a video of this post at the bottom if you’d rather watch me work.
I started a new project, mostly as an experiment to help me practice with some technologies, with the idea that this will become useful at some point in time. In my case, I created a new data model for an event registration system. We could really use one at SQL Saturday/Day of Data and right now everyone that runs an event has to handle that themselves.
In any case, here’s my basic starting model. You don’t need to look at the details, just know I started here by creating a model from scratch in Redgate Data Modeler. I have 7 tables that I setup.
I want to get this model deployed into a database. I could export the SQL, which is fairly easy, but let’s make this easy with SQL Compare 16. I upgraded after the new release, and I want to see how this works.
Data Modeler Integration
In SQL Compare we have a new source for connecting to database code. In this case, it is Redgate Data Modeler Integration. That doc pages describes it, but when I open the tool, I can see a new choice in the connection drop down.
This is a preview feature for now, but it does work. Once I select this option, I see the settings change. In this case, I need an API to get started.
The documentation says to go to your Settings page in Redgate Data Modeler. When I went there, I got an API section at the top, asking me to enable API access (which I clicked before taking this screenshot) and also a set of Notification Settings below this.
Once I clicked things, I had the API token listed. I’ve blacked out most of it here, but I clicked the “copy” button to save this to the clipboard.
I then pasted this into SQL Compare and the model drop down populated. I could see both my models listed.
For the target, I connected to an empty database that I created with this code:
CREATE DATABASE SQLSatRegistration_1_Dev
I could see my comparison looked correct, so I was ready to see if I could deploy my model.
I ran the comparison and as expected, it sees everything as new in the model and nothing in the database. I selected all objects and then clicked Deploy.
Since this is the first time I’ve deployed the model, I decided to let SQL Compare do the work.
After looking at the script briefly, I clicked Next and got the final, do you want to deploy, screen. I clicked Deploy here.
I had selected the recompare after deployment option, so once this finished, I saw this screen. Everything is the same on both sides of the comparison.
When I checked my database, I see the objects there.
Summary
This was a short look at taking a new model I’m working on and deploying it to a database for some actual work. I took a model previously created in Redgate Data Modeler and using SQL Compare, I deployed all the objects to a live database.
This is a great integration and it’s much smoother than saving a script and then opening that to run it. I like tools working together, and I could see this being handy for architects and modelers that want to quickly test out their changes.
SQL Compare is an amazing tool that millions of users have enjoyed for 25 years. If you’ve never tried it, give it an eval today and see what you think. Give Redgate Data Modeler a try and see if it helps you and your team get a handle on your database.
![2026-01_0173_thumb[1]_thumb](/ssh-https-voiceofthedba.com/wp-content/uploads/2026/01/2026-01_0173_thumb1_thumb-1.png "2026-01_0173_thumb[1]_thumb")
![2026-01_0172_thumb[1]_thumb](/ssh-https-voiceofthedba.com/wp-content/uploads/2026/01/2026-01_0172_thumb1_thumb-1.png "2026-01_0172_thumb[1]_thumb")
Voice of the DBA 
