David Morgan-Gumm (00:00)
Welcome everyone to the fourth episode of the SQL Squared podcast where I am joined by Lee Brownhill. Welcome Lee.

Lee Brownhill (00:08)
Bye, y'all.

David Morgan-Gumm (00:09)
it's great to catch up because I know that we've both been on the finding data friends podcast recently. Um, uh, great timing. Mine was just released yesterday. Oh, thank you. Thank you. Um, yeah, it, uh, I had, you know, Ben and Jess were saying it's, one of the most fun ones they've had in a while. Um, it kind of went a bit, a little bit off track.

Lee Brownhill (00:13)
Yeah.

And it was great, by the way.

Yeah.

Exactly, the rule works.

David Morgan-Gumm (00:33)
I think some raw eggs or something were mentioned halfway through. I don't know what happened. But yeah, I

kind of want to know what the experience is like for you because I know your episode hasn't released yet, but it's been recorded.

Lee Brownhill (00:43)
Yeah, yeah, it was great fun. ⁓ horror because they're super warm, super welcoming, super friendly. And I've met Jess a few times now at various events. I think most recently at the Red Gate Summit London. And I've always been impressed by just how knowledgeable she is just on the fly, you know, when you can tell that it's genuine knowledge and people can just throw questions at Jess and she has

David Morgan-Gumm (00:51)
Yeah, yeah, yeah.

Lee Brownhill (01:12)
and well thought out, experienced answer every time. ⁓ But yeah, it was great fun. They were both lovely, both really helpful and friendly and welcoming. So yeah, good times.

David Morgan-Gumm (01:23)
Yeah, I had the exact same experience. ⁓ It was a nice thing to be on. It was kind of good that it was only 15 minutes. ⁓

Lee Brownhill (01:30)
Yeah, it was one of the attractive things to me because I'm quite new to podcasting and being on shows such as your own. So it was a nice entry sort of piece to do because of the it was such a short episode, right?

David Morgan-Gumm (01:46)
Yeah, I agree.

I think that's kind of what it's based around, it? They're trying to introduce people to the community. That's the whole point in it.

Lee Brownhill (01:54)
Yeah, at any show

where I can talk about SQL Server and food together, I mean, I'm there, That's a secret.

David Morgan-Gumm (01:58)
I'm not gonna ask you what your favorite food is. We'll leave that for the for when the Finding Data Friends podcast comes

out. ⁓ I suppose what did you enjoy most about being on the show?

Lee Brownhill (02:10)
So the thing in which, yeah, I mean obviously data and food together, it's a match made in heaven. And it's the first public thing I've done in that kind of thing. it was very much a learning curve and pushing myself out to my comfort zones and things like that. it was great fun. And like I said, I have a relationship of sorts with Jess anyway.

David Morgan-Gumm (02:16)
Yeah.

Lee Brownhill (02:38)
I've met her a few times now at various events around the UK and always respected her work regarding DBA tools, sort of thing. So to actually get to do something like that with both Ben and Jess, yeah, great fun. Really enjoyed it.

David Morgan-Gumm (02:47)
Yeah.

I hope to have them on this podcast one day. Yeah, I've seen, I haven't gone and kind of talked to them in person at conferences, but I've been to their talks and stuff and I know that both of them speak, you know, in a lot of depth and they know what they're doing.

Lee Brownhill (02:57)
That'd be amazing, yeah.

Yeah

For sure, for sure. I'm not so up on Ben's presentations because I haven't been to any events that he's actually been into the UK yet, but I have seen his stuff online and yeah, always good times with Ben.

David Morgan-Gumm (03:28)
Yeah, yeah, it makes it, it makes it fun. Right. Well, one of the obviously SQL squared podcast focused on the guest focused on you, your names in the title. It'll be on the thumbnail as well. I kind of want to know a little bit about

Lee Brownhill (03:30)
Yeah, very much so.

Thank

David Morgan-Gumm (03:46)
your career i know like you know you've been in this line of work for quite a long time can you tell me a little bit about your career journey and kind of what sparked your interest in data in the first place

Lee Brownhill (03:57)
Of course, yeah. So many moons ago, more years than I'd care to admit at this point, I started off in a general sort of IT support role. ⁓ Classic sort of entry point into the IT world, tech world, et cetera. So jack of all trades, pretty much master of none. And the business I was with at my very first sort of proper IT job.

David Morgan-Gumm (04:02)
Yeah.

Mm-hmm.

Lee Brownhill (04:22)
They migrated from a Unix based ERP system to Dynamics NAB at the time, which was obviously powered by backend SQL Server. And I got really interested in that particular server. And it was like this scary magic box to me and most other people in the business. So I started to look into that magic box essentially. And the more I worked with that SQL Server, the more I knew I wanted to actually just

do that rather than all the other parts of my job because I've never had much interest in various aspects of IT. But the one thing I have always had interest in is the data and specifically SQL Server. So after a lot of training on my own part, I eventually got a production DBA job at a SaaS company. So traditional DBA job, keeping the lights on, keeping things running, making sure your backups are in good state of play and things like that.

David Morgan-Gumm (05:15)
Mm-hmm.

Lee Brownhill (05:20)
And then fast forward several years, I ultimately ended up as a kind of a mixed role, like a developer DBA kind of role. So I was working on a hybrid sort of small team, ⁓ all centered around making that particular application scale. Because originally it was quite a small customer base. And that company got purchased by a much larger US version. And all of a sudden we had to make that database and that product scale to

the kind of scale that Fortune 100 customers put on it, which is a times 10 difference really. So that was one of my most favorite roles in hindsight because that's where I learned the most because I had to almost become the glue between developers and your infrastructure DBAs and get them speaking together in the same language almost because...

David Morgan-Gumm (05:59)
Yeah.

Lee Brownhill (06:17)
They obviously coming from very different areas in terms of their priorities, their wants, their needs. So

David Morgan-Gumm (06:21)
Yeah.

Lee Brownhill (06:23)
actually needs the conduit between those two teams was actually really challenging, but really fun as well because ⁓ we had to do some drastic work on that product to actually get it to scale.

David Morgan-Gumm (06:35)
I think you find similar stories in a lot of companies that have much smaller teams. ⁓ You know, the, experience is mostly with small teams as well. ⁓ And, you know, you learn a lot in those environments because you need to pick up a lot and you need to be a hybrid role. Like you don't have the, you don't have the people.

Lee Brownhill (06:38)
Yo

So.

David Morgan-Gumm (06:54)
to take the individual roles and move that forward. You have to all work together and all muck in. ⁓ And it's a great environment to be in.

Lee Brownhill (06:57)
Yep. Totally. Exactly that, yeah. And it was

just one of those really challenging, but in hindsight, rewarding roles because I had to, yeah, just really get the differing teams all working towards the same goal, which was all about scale, right? So that could range from holding...

sort of show and tell discussions and sessions between the more junior devs about what we need to stop doing within code, which doesn't scale from that sort of thing to more complex things about general architecture and how we need to actually look into things. And we actually ended up moving to one of your favorite topics. So I know from the finding data friends, which was managed instances, we actually moved to managed instances for that particular task. So that came with its own.

challenges and benefits also as well. But yeah, it was a really rewarding time. And now, yeah, up until more recent experience, now I work as a more of a consultant slash DBA. So we advise and help our clients to get the most out of their workloads within Azure and AWS. And that's really been rewarding from my point of view because I've

suddenly got hands-on experience with AWS, which is always a bit of a mission on my experience, so to speak. It had never actually worked. It had all been within Azure up until this point.

David Morgan-Gumm (08:21)
Mm-hmm.

How do you find AWS compared to Azure?

Lee Brownhill (08:33)
Um, it's, it's got its benefits and its negatives compared to Azure. In my opinion, there are things it does better. There are things that are does not so great as Azure in my opinion, but again, it's the right tool for the right job at the right time, right? Um, but now I'm in my current role, I'm surrounded by real seasoned experts of the field in my current role. So it's quite humbling and exciting times for me because I'm constantly learning new things. Um,

David Morgan-Gumm (08:37)
Yeah.

Mm-hmm.

Yeah, of course. Yeah.

Mm-hmm.

Lee Brownhill (09:02)
and there's plenty to do always. yeah, ⁓ good times, good times, really enjoying it.

David Morgan-Gumm (09:09)
That's good. I suppose as more of kind of a DBA consultant, I can imagine that you're very focused on performance and efficiency and getting the most out of the least amount of money. Let's put it that way.

Lee Brownhill (09:18)
Yes.

Exactly,

right? Exactly, because there are always budgets to adhere to. It's not just a case of turning sliders up indefinitely. These businesses have to remain efficient in every sense of the way. So we're there to guide them and help them through that journey because some are new to the cloud. Some of their been there for a while, but need ⁓ that extra expertise to actually understand.

David Morgan-Gumm (09:33)
and

Mm-hmm.

Lee Brownhill (09:50)
how to get the most of their cloud environments and things like that. So ⁓ it's real fun, real challenging and every day is different at the moment. So it's really, really exciting.

David Morgan-Gumm (09:52)
Mm-hmm.

Amazing. Yeah. Sounds exciting. And how I'd love to just turn the knob up to the max and then just leave it there and not have to worry about the cost. Yeah. Especially in the cloud. All right. Thanks for that. That was a really good insight. ⁓ know, extensive career, ⁓ extensive expertise, but it's always good.

Lee Brownhill (10:03)
Yeah it is.

Yeah

.

David Morgan-Gumm (10:24)
So even after you've got, was it 15 plus years that you've been working in the industry? Nearly 20. Yes.

Lee Brownhill (10:31)
somewhere around that. don't like to talk about numbers too much, David. But yes, ⁓ I've, I've worn many different hats around SQL server now. So, ⁓

it's, I feel like I've got some, ⁓ good experience to actually start giving back to the community, which I'm very keen on doing this year.

David Morgan-Gumm (10:48)
Yeah.

Yeah. And, you know, constantly learning as well still, which is really good to see.

Lee Brownhill (10:56)
Exactly.

The day I'm in a position where I'm not learning is quite troublesome for me and quite demotivating because I've always had this thing about I never want to be the smartest person in the room about a topic, right? Because there's just no growth there at all ever. And the second I do find myself in that situation, I will instantly make myself uncomfortable by going in the situation where

David Morgan-Gumm (11:06)
Mm-hmm.

Mm-hmm.

Lee Brownhill (11:26)
I'm surrounded by very intelligent, very clever, very seasoned experts. That's where I make my best, I think.

David Morgan-Gumm (11:33)
If you're the smartest person in the room, you're in the wrong room. I've been working on recently a mentor of mine was saying to create a bit of a T-shaped skill profile.

Lee Brownhill (11:35)
Exactly right.

Yes.

David Morgan-Gumm (11:48)
⁓

you know, make sure that you've got a foundational amount of knowledge and expertise over a large range of things. ⁓ but make sure then you, ⁓ focus on a small subset of that and become an expert in that field. And kind of that's how it's one of the main ways people, ⁓ become quite successful in what they do. ⁓ you know, the goal, the goal of this podcast is to, I suppose, help people organically learn what is involved in

Lee Brownhill (12:05)
Yeah, I totally agree.

David Morgan-Gumm (12:17)
data development and SQL in particular, ⁓ and to kind of help them make the right decisions along their journey and to have that start to that journey as well.

Lee Brownhill (12:28)
Yeah,

for sure. I think the going back, I've heard of the T shaped skill sets as well, because I think one driver for that is the cloud ultimately. And yet the cloud is obviously not a new thing now, but you can no longer just be a data person that just opens up SSMS all day and that's you, right? Those days are gone. You have to know a bit about networking. You have to know a bit about security.

David Morgan-Gumm (12:51)
That's right, yeah.

Lee Brownhill (12:54)
you've got to know these other things to a certain degree. You might not be a master of them, you've

got to have an understanding and an awareness of them for sure.

David Morgan-Gumm (13:03)
I mean, if you're just a SQL developer working on on-prem solutions and you end up moving to the cloud, starting to use the cloud, I feel like even just opening like the Azure portal, you just get provided with a ton of stuff that you don't, one, don't necessarily, you might not necessarily care about it, but clearly the fact that it's provided and then it needs to be managed, you need to know about it. It's all part of the fun.

Lee Brownhill (13:21)
Hahaha

Yeah, exactly.

Exactly.

David Morgan-Gumm (13:33)
Okay. T-SQL T, I've already mentioned it's something that I, I'm definitely keen on learning more about myself. I don't really implement automated SQL tests in my own environments. Kind of the more, more integration tests, I suppose, on some of our surrounding systems and within our kind of data pipelines and things like that to make sure that, you know, the integrations

Lee Brownhill (13:35)
Yeah.

you

Good.

David Morgan-Gumm (13:59)
run well rather than the systems within SQL and the processes. What I've done in the past is kind of, I suppose, alerting processes within procedures and things like that to check the data as it goes along rather than ⁓ using kind of like a unit testing framework. So can you just give a brief introduction to T-SQL-T for those that don't know much about it?

Lee Brownhill (14:00)
Mm-hmm.

Mm-hmm.

Of course, yeah, sure. So T-SQL T is ultimately an open source framework ⁓ to enable sort of writing T-SQL tests or unique tests, sorry, within T-SQL. So you're not having to learn a separate program language. You can just use T-SQL to write these tests. And in my opinion, it's a fantastic tool. It saved me quite a few times in previous roles in terms of highlighting edge cases that...

you just wouldn't have found regarding normal smoke testing. So it's really, ⁓ I'm a huge convert. ⁓ I used to use T-SQL-T quite heavily in a previous role, couple of roles ago now. There's a possibility I might need to start using in the future. So what I thought I'd do is start blogging about it ultimately. So I could refresh my memory around T-SQL-T, create some blog posts, ⁓ and sort of document my learning ultimately.

Because I do I'm a firm believer of the old adage that you know the best way to learn is to teach something It's a bit trite, but it's true And I truly believe it so yeah, I mean t-sequency I'm a huge convert, and it's really helped us out in previous roles

David Morgan-Gumm (15:42)
That's good. Yeah. And I definitely agree with the best way to learn is to teach something. I've been doing that a lot recently with AI and coding agents. I've been doing a lot of research in my, you know, in my spare time.

Lee Brownhill (15:46)
Yeah.

Okay. Yeah.

David Morgan-Gumm (15:57)
⁓ to help basically help improve my team's workflows ⁓ now and moving forward into the future. And it's one of the things that really sparked ⁓ the idea behind SQL Squared. know this is SQL Squared. The name is meant to signify that it's much more than SQL. Data is much more than SQL.

Lee Brownhill (16:19)
Mm-hmm.

Yeah.

David Morgan-Gumm (16:22)
and we talk around everything that surrounds that, know, and t-sql-t being one those things today. ⁓

Lee Brownhill (16:28)
Yeah, I mean it's not loved by all and I think there are various reasons for that. So I think there's a big investment at first because

ultimately every procedure you want covering unit test for every procedure or every function in your system eventually to get the true benefit from it. So it's quite a heavy investment initially because obviously you'll be starting from scratch.

David Morgan-Gumm (16:44)
Mm-hmm. Mm-hmm.

Lee Brownhill (16:56)
But there are also tools you can use to sort of help you along that way and streamline that initial sort of input required. So yeah, I mean,

David Morgan-Gumm (17:05)
Mm-hmm.

Lee Brownhill (17:06)
there are certainly the naysayers around T-Security and I understand their points, but from my experience, it's been a really great, almost security blanket around you sometimes because you will just be coding away, making a change to a sort procedure or a new piece of functionality for your product, et cetera.

David Morgan-Gumm (17:16)
Mm-hmm.

Lee Brownhill (17:26)
And then you think, yep, we're all good. And all of sudden, some unit tests will start failing. And I always think back to almost, it was kind of my mentor in that role, really. And he'd written one of these unit tests. And I actually had the audacity to go to him saying, one of your unit tests is incorrect. And lo and behold, it was my code that was incorrect. But that was one of the, that was the, what the time that really

David Morgan-Gumm (17:45)
Yeah,

Lee Brownhill (17:54)
converted me really because I would never have found this edge case bug ⁓ and it's best we find internally before you know, obviously the client or the customers do.

David Morgan-Gumm (17:55)
yeah. Well, yeah, that's the whole point in unit tests, isn't it? no, yeah. Yeah, definitely.

Lee Brownhill (18:05)
Exactly. I mean, and it's something that I don't see often within the SQL world being used.

I mean, your .netters and et cetera, it's hand in hand, right? They're always writing associated unit tests for any of their code. It's just, certainly the ones I've worked with, you know, it's just, it's just one thing they do, but it's a very new thing, still and rare thing I've found in, ⁓ the SQL side of things.

David Morgan-Gumm (18:16)
and

Yeah, yeah. it's one of the major improvements that I've actually seen from this whole AI agent stuff that I've been working on. It can write unit tests quite well. One of the improvements that I've seen in my workflow is the efficiency gain in writing unit tests is

Lee Brownhill (18:36)
But.

Mm-hmm.

David Morgan-Gumm (18:54)
astronomical and I've been testing it with t-sql-t as well ⁓ in the background this month. Obviously you have to give it quite a lot of direction to know what you want to get out of it but it can write ⁓ large suites of tests across an entire database, saving you days of time. Obviously it needs to always be checked and then tested itself ⁓ because

Lee Brownhill (18:58)
Okay, cool.

cool, okay.

Yeah.

David Morgan-Gumm (19:21)
Ultimately, you know how the system works and the AI doesn't. But the more instruction you give it and the more context you provide it, the better it'll do. I've been working with ⁓ developed my own MCP this month. Used a, have you seen this? There's quite a good open source community for MCPs now.

Lee Brownhill (19:37)
Okay.

Right, okay.

David Morgan-Gumm (19:46)
Model

context protocols, I should explain what that is. They're they're applications that sit between your AI agent and whatever other system you're working with to give the AI agent much more context over exactly what you're doing, specifically what you're working with. So it's not using a whole, you know, just general archive of stuff to know what it's doing. And you can do it to SQL, SQL Server.

Lee Brownhill (19:58)
Okay.

Right. Yeah.

David Morgan-Gumm (20:15)
So I've been testing internally, basically writing a ⁓ .NET ⁓ based MCP server to run on Visual Studio Code, which sits within the agent. And it allows the agent to actually go and query your database or your data warehouse ⁓ and to figure out how the data pieces together to then understand what the flow of the data is and well,

Lee Brownhill (20:24)
No, I'm okay.

Okay, cool.

David Morgan-Gumm (20:43)
what I've seen success in writing unit tests in that regard, but also diagrams and documentation as well of your procedures and environment. I'd recommend having a look into it because it's been, yeah, it's been really powerful, really, powerful. Sticking with T-SQL-T, and I know you've mentioned, you just kind of mentioned it at a high level that unit tests are hard to implement.

Lee Brownhill (20:51)
Yeah, okay.

Yes, I will do. That sounds cool.

David Morgan-Gumm (21:12)
at the start and I totally agree with you, which is one of the reasons I've been looking at trying to implement all the AI stuff in here as well. But in general, once you've got T-SQL T-setup or whatever unit testing process you're using, are there some common challenges that people face that you could, when adopting these technologies, that you can...

Lee Brownhill (21:18)
Yeah.

David Morgan-Gumm (21:41)
help them overcome.

Lee Brownhill (21:43)
Yeah, initially, it's an investment and you can't get around that. Whether that's, I mean, that's changing now a little bit based on what some of the things you've just spoke to, but ultimately, at first, it feels like yet another unnecessary task to do to get that Jira ticket over the line, right? Because in that particular role where I initially had the experience with it, you couldn't get your...

David Morgan-Gumm (22:00)
Yeah.

Lee Brownhill (22:09)
you couldn't ultimately get your PR even approved and merged in until you had a associated unit test there. And it was all integrated within our CI CD pipelines. So you had to have associated unit tests. And more importantly, they all had to pass before they could, they would even allow the PR to get merged in. initially, in my experience, development teams are normally super busy, super overburdened already.

David Morgan-Gumm (22:21)
Mm-hmm. Yeah.

Mm-hmm.

Lee Brownhill (22:39)
and getting pushed to deliver things as soon as possible. So the introduction of trying to suggest there's more things to do can sometimes be a challenging ⁓ thing to announce. However, it's worth it, I believe, ultimately in the long run. And you can obviously take it small at first, piecemeal it and just have smaller pieces of work every time you do it. But yeah, the ultimate thing for me

was actually introducing that to the development teams and actually getting them on board because they're busy already, right? They're really busy and overburdened. yeah, that was the largest, the biggest challenge I think I faced at the time.

David Morgan-Gumm (23:12)
Mm hmm. Yeah.

That's really good. That's really good to mention. My experience has actually been in the opposite direction. So I've tried to implement testing in many, many, many different forms ⁓ over the years.

Lee Brownhill (23:28)
Okay.

David Morgan-Gumm (23:37)
I've gotten much more resistance from leadership rather than developers. I think I actually got a quote one time that somebody said of, you know, why do we need to test? It should just be right first time. But I think that's, you know, that's clearly the ⁓ whole thing of it. It's rarely right first time. And it's better to test and make sure that

Lee Brownhill (23:43)
interesting.

Yeah.

David Morgan-Gumm (24:06)
what you've written works as expected and to push it live and figure out that something's gone wrong and cause downtime in some area.

Lee Brownhill (24:08)
Exactly though.

And the problem only becomes

more exacerbated, right? The larger your development teams as well, because you've got different people touching different bits of code, sometimes overlapping. having that, ⁓ yeah, that security blanket overlooking everything to make sure that the core functionality of those procedures or functions still remain intact, ⁓ that's priceless. And it proved itself so many times.

David Morgan-Gumm (24:16)
Yes. Yeah. Mm-hmm. Yeah. Yeah. Even if it slows things down.

Yeah. Even if it slows time to delivery, it's that security blanket is so worth it.

Lee Brownhill (24:46)
Yeah, and it was challenging. not going to, I can't deny it was challenging initially because we were working in two weekly sprints. So every two weeks, big release of a big release, big update out. So every two weeks there, there was this cadence that had to be met. We had to deliver certain features every two weeks and things also have to remain functional and performant and secure.

David Morgan-Gumm (24:46)
I ⁓

I do the same. Two weeks. Mm-hmm.

Lee Brownhill (25:13)
So it's really difficult balancing all those things at times together.

David Morgan-Gumm (25:17)
Mm-hmm. ⁓ I agree.

And I know we're going to come onto Redgate in a second, but I know that Redgate have a test data manager tool as well, don't they? Have you got any experience with that?

Lee Brownhill (25:26)
They do, yeah.

I've used the RedGate SQL test tool, which integrates beautifully with the T-SQLT framework. And I've actually got some, the last blog post on the T-SQLT did actually cover a bit of that. But that tool is fantastic for automating some of the tasks and giving you a nice graphical user interface for actually seeing.

David Morgan-Gumm (25:37)
Yeah. Yeah.

Lee Brownhill (25:54)
⁓ how many of your tests are passing, the percentage they're passing, the ratio, and more importantly,

David Morgan-Gumm (26:00)
well yeah, I never really thought about that. Mm. Mm-hmm.

Lee Brownhill (26:00)
one of my favorite features of that tool is the code coverage. It shows you what code coverage you've got within each of your objects. within your procedure, can actually, it highlights the areas that are covered and the amount of time you think, yeah, I've got that, ship it, it's done. You go and look at your code coverage and you've only got two thirds of the code covered.

That's, ⁓ I love that feature of it because it's just so easy, right? It's there.

David Morgan-Gumm (26:26)
Okay. Yeah, yeah, yeah, yeah. I need to look into it again.

I ⁓ played around with it a couple of years ago, but never kind of put anything into production or brought it through. All right then. Yes, we're talking about Red Gate. congratulations on becoming a community ambassador. ⁓ I suppose, what does being a community ambassador mean to you?

Lee Brownhill (26:34)
Thank

Thank you.

⁓

I think firstly it's yet another platform for me to be able to interact and share ideas and share things with the community. I think it's a great avenue to be able to do that and it's you know in the grand scheme of things it's another platform for me to be able to speak to people and make new connections, make new data friends and yeah.

David Morgan-Gumm (27:12)
Yeah. Yeah.

Lee Brownhill (27:20)
just all of that stuff really. I'm so excited and it was a real surprise and a real treat. But I've been quite vocal in my love for Redgate tools over the years because it's been such a time saver for me. over the years, whether that's SQL prompt, whether that's SQL Monitor monitoring all your databases, whether that's Redgate SQL tests, it's just been...

David Morgan-Gumm (27:33)
Mm-hmm.

Lee Brownhill (27:48)
I love their tools and I think if you are, if you've worked in

SQL or databases for a while now, you've had some exposure to their tools and there's a reason for that, right? Yeah.

David Morgan-Gumm (27:56)
Yeah, yeah, yeah, yeah, yeah. You know who Red Gate are. Yeah.

I just think that even the Red Gate community in itself, which obviously is kind of a subset of the data community, is thriving. It's great. They do a good job over there. And every time I see Red Gate at conference or an event, everybody's always kind of flocked around their booth. They're always really, really friendly guys.

Lee Brownhill (28:09)
Yeah

David Morgan-Gumm (28:19)
from my experience with working with them recently, like so supportive as well.

Lee Brownhill (28:23)
Yeah, very much so. ⁓ Great, great company. And yeah, I think they're, they're advocates as well. So you're Steve Jones, you're Kellan Gorman, you're Grant Fritchie. They're amazing. And I, I will always try and attend their talks if I'm at a session they're speaking because there's a reason they've been chosen for those roles. And it shows, right? It shows because they've got, they've got such a vast amount of experience. It's incredible.

David Morgan-Gumm (28:32)
Mm. Mm-hm. Mm-hm. Yeah, yeah,

Yeah. How do you see yourself contributing?

Lee Brownhill (28:54)
Um, well, firstly, um, I think I'm going to be able to, yeah, leverage that platform and whether that's blogging. So, um, I'm to be able to, on my, on my own platform, I'm going to be able to show how some of their tools have actually helped me over the years. And rather than just saying these tools are great, actually give examples of how, yeah, exactly. I mean, there are aspects of.

David Morgan-Gumm (29:17)
Reasons why in examples and what people through it. Yeah.

Lee Brownhill (29:23)
Redgate SQL prompt, example, that ⁓ I use now. And if I don't have that tool available to use, I feel almost lost because you get that used to using these shortcuts and various aspects. So yeah, the blogging is going to certainly contribute to that. The events and conferences, Redgate Summit conferences are some of my favorites now just for the diverse amount of topics.

they're on show, but obviously all centered around the data platform. So yeah, and continue to build that online presence ultimately. So it was a real surprise and a real treat, really excited about it.

David Morgan-Gumm (29:52)
Mm-hmm.

Good, good,

good. I've listened to a few of their Simple Talk episodes as well. Kind of got a bit of a similar format to this. Hopefully, I hope to be on that one day, I think, and try and get other community ambassadors involved in SQL Squared. think a really good partnership.

Lee Brownhill (30:17)
Yeah.

David Morgan-Gumm (30:22)
to work together, something that I'm definitely working towards. Yeah, you'll take it, you'll take it. I'd take it as well. I suppose for the audience out there, you know,

Lee Brownhill (30:23)
Yeah. And quite frankly, anywhere I can be on list on the same sort of, you know, alongside the likes of Aaron Bertrand and Ola Hallangren, then, you know, I'm happy. ⁓ I don't know if I've just deserved, but I'm taking it.

David Morgan-Gumm (30:51)
Lee and I have just talked about a few of the Redgate tools and the Redgate community in general. Totally advocate what they're doing and I'll put a few links to their blog and podcasts in the description below. And if you have any questions or want to find out more or want to just connect with me and Lee, please message me at mailbag at SQL squared dot co dot UK.

and I'm more than happy to answer your questions and have a chat and get in touch and link you with people to learn more and other resources, especially Lee's blog on T-SQLT at the moment.

Lee Brownhill (31:29)
Yeah.

David Morgan-Gumm (31:31)
So giving back to the community then, we've mentioned a couple of bits already about what you want to do, working with the Red Gate community and what your goals are for this year. I know very specifically we've got sequel bits happening in June. Unfortunately, I am gutted that I can't go, ⁓ but also very ⁓ excited to be in Bordeaux for the wine festival. So that's why.

Lee Brownhill (31:49)
no.

⁓ okay. That's a very strong

second place, yeah. Yeah.

David Morgan-Gumm (31:59)
Yeah, yeah, yeah. I love SQL bits as a conference. It's great.

And it's an amazing place to network, find out about new tools and just learn a load of new things. I always walk away, you know, learning a ton from that conference. Normally I go to the free day, but last year I was lucky enough to go to one of the paydays, paid days as well.

But yeah, some of the talks are in depth. I know they've got the agenda out at the moment, but I haven't looked through it because obviously I know that I can't go. But you mentioned that you're a SQLBits helper this year.

Lee Brownhill (32:39)
I am, and I'm very excited. It's the first time I've ever done it. So it's going to be an experience for sure. I almost was a helper in 2024, but I got ill literally the day before. So it was, I was gutted. this year, yeah, exactly. So this year I'm super excited to be a helper and wear one of those famous orange t-shirts and you know.

David Morgan-Gumm (32:53)
Mm-hmm. ⁓ So you didn't end up going to the conference at all. ⁓

Lee Brownhill (33:09)
helping out and moderating some sessions. I'm

really excited. I'm going for two days as a helper. I have been to bits numerous times and I've also attended one of the paid training days before. I attended one of Eric Darling's SQL performance days in Manchester. I can't remember what year it was in Manchester, but a few years ago, it might have been 2019 actually, all the way back then. So, and that was fantastic, obviously.

David Morgan-Gumm (33:11)
Right.

Lee Brownhill (33:38)
⁓ and hurt my head a lot of the topics

that have been discussed. But yeah, can't wait, really excited. And it's just another one of those avenues about really trying to give back to a community that's just supported me and given me so much over the years.

David Morgan-Gumm (33:50)
Yeah, yeah,

Being a helper at a conference is one of the best ways to do it. Like it's such a, suppose I always see people doing it and it's kind of such a selfless thing that they're doing because, you know, go into it as a helper in certain sessions. I assume you get provided sessions.

Lee Brownhill (34:05)
Yep.

Yep. Good question, because I never actually got that far on my previous attempt to be a helper.

David Morgan-Gumm (34:11)
rather than choosing sessions? Or do you get to choose sessions as a helper?

All right, so I'm guessing

it. So you haven't done it for this year yet, so it probably happens when you get to the conference.

Lee Brownhill (34:23)
No, so I

believe that the training days, those are allocated at the time, sort of on the day, depending on how much resource they've got available and all those sort of factors they have to consider. So we've got some catch-up calls soon regarding the helper team. So I'm sure we'll find out more information then. So who knows?

David Morgan-Gumm (34:31)
OK. Fair enough. ⁓

Great. Yeah, because I can imagine a thought about trying to do it in the past, but going to the talks that I want to go to is the reason, I suppose, that I go to the conference. ⁓

Lee Brownhill (34:54)
Yeah, yeah, yeah, exactly.

David Morgan-Gumm (34:58)
I think I might do it sometime in the future, definitely, you know, want to want to continue, want to continue learning, to continue discovering new products, like I said about the, AI agent stuff earlier. I just love working with the newest stuff, the newest tech, and I, and I'm always trying to bring it into my workflow and change things. There's a, I might have mentioned it on the podcast before, I think my favorite word is innovation.

And I just can't, I'm always trying to innovate and you know, sometimes you get change fatigue with it if you do it too much. And sometimes especially in the world of business, you've got to be careful about the things that you innovate on and choose wisely what you're trying to do.

Lee Brownhill (35:37)
Yeah, for sure.

Of course, Yeah.

David Morgan-Gumm (35:48)
⁓ recently actually had a process where we've been doing bit of a migration of a suite of tools and get the opportunity to change and innovate on things during a migration process. And we actually had we had

people really bought in, we had a load of ideas of things to change and improvements to make, but we didn't have a lot of time to do the migration and lot of resource and so we had to kind of balance what change we made. they can be, yeah.

Lee Brownhill (36:13)
Mm-hmm.

Right. It's exact. That's the key word. That's what was screaming in my head then when you're talking about it's a balance, right? Because you can leverage new tools, new processes to help you along this journey, but also there's a debt to that as well, right? So

sometimes the biggest thing you've got to adhere to is a timeline with, in my experience, with these sort of larger projects. And there's always a deadline.

David Morgan-Gumm (36:37)
Yeah.

Lee Brownhill (36:44)
And there probably should be really, because they could go on forever. So it's a balance, because I've been in similar situations where working in agile environments, you do have to be open to flipping the script a bit and coming at things from a completely different angle. have to sometimes just to get out of the line.

David Morgan-Gumm (37:04)
Yeah. Yeah. Yeah.

Lee Brownhill (37:07)
But also, yeah, there's a risk involved there. yeah, balance for sure.

David Morgan-Gumm (37:12)
I've definitely, I've come across integrations and processes within data warehouses in the past where people clearly didn't take that step and they built new processes with new applications using old kind of models and you

Lee Brownhill (37:35)
Yeah.

David Morgan-Gumm (37:35)
ways of working. And you end up not integrating with stuff correctly. One of the examples I've got is the difference between soap APIs and REST APIs. People developing, unfortunately, the other way around, developing as

Lee Brownhill (37:41)
Exactly.

Okay.

But.

Okay.

David Morgan-Gumm (37:59)
on a soap process rather than a rest process, but they were very used to the rest process. And the way that you interface with soap APIs can be very, different. And, you know, if you use the the same philosophy, I suppose, and the same way of developing interfaces with those types of APIs,

Lee Brownhill (38:04)
Okay.

Yeah,

yeah.

David Morgan-Gumm (38:27)
Doesn't, it doesn't necessarily work. ⁓ even, even though they're theoretically the same thing, you know, you are just interfacing with stuff and, and. ⁓

Lee Brownhill (38:29)
It's so true. It's so true.

Yeah. And you've made me think of something from my experience and my world in terms of migrating clients and getting workloads into the cloud. A common pattern I've seen is something I've termed in the past of on-premise in the cloud, where because it's a known quantity, people want...

There's this...

often a want to take those same functionality, the same technologies, the same constructs of those technologies and just plump them into the cloud. And there you go, which is obviously getting you into the cloud, but you're not making the most of those cloud native technologies. And you are instantly limiting yourself from what you can actually do in the cloud. But again, it's a balance. ⁓

David Morgan-Gumm (39:00)
Yeah. It has to act as a bridge. Yeah,

I said it has to has to act as a bridge, I suppose.

Lee Brownhill (39:27)
Yep,

exactly.

David Morgan-Gumm (39:28)
Like it's

one of the reasons I loved managed instance. ⁓ I found it so much easier to migrate from an old kind of VM hosted on-premise solution to a managed instance, which obviously is a managed service within Azure. And there were changes obviously that needed to get made as part of that migration process. However, it was more seamless.

Lee Brownhill (39:42)
Yes.

David Morgan-Gumm (39:54)
than using other technologies. And it allowed us to move into that managed environment rather than having on premise in the cloud. Yeah. Yeah. Yeah.

Lee Brownhill (40:01)
Yeah. I mean, definitely plus one for managed instances. I know we spoke about this before, but I'm a huge fan and we're actually early adopters at the time when their fares came out and we had a lot of problems with them because in those days things were changing on the daily basis. So

I remember when you first went up a managed instance, you couldn't even choose the default collation on them. You were just given a specific collation. So then we had to get our development team to remove all these hard coded.

I know, don't shoot me, but there were hard-coded collations within our code.

David Morgan-Gumm (40:32)
Yeah.

Lee Brownhill (40:34)
We had to get them to change all of those. And then two weeks later, all of a sudden you could change a collation. that was technically wasted effort from the development teams, but that's just the price you pay being on NewTek, right?

David Morgan-Gumm (40:46)
Yeah.

Yeah, that's one of the downsides of it. mean, we have not even tried to adopt Fabric yet. Not even really considering it at this point. Yeah. I hear too many bad things. No, I I...

Lee Brownhill (40:54)
Okay.

It's a hot potato at the moment, right? I'm going to not say much here David about fabric because I don't know.

David Morgan-Gumm (41:08)
I'm actually, I'm going ⁓ 10th of June, ANS have a fabric in a day workshop on. So I'm taking a few of my members of my team to that. It's just a, it was literally a free event. ⁓ I was speaking to our Microsoft account managers and they put us onto it. they're Microsoft are paying for it, which is, which is great. So hopefully, you know, next, next podcast, I do, I'll have a little bit more knowledge in the fabric space, but I just, hear.

Lee Brownhill (41:14)
cool, okay.

Yep.

Clear? Clear?

Yeah.

David Morgan-Gumm (41:36)
So many bad things about it. was actually a post that Brent Ozar put up on LinkedIn a few weeks ago. The one that was all that everything was down. he was just in a one-liner, I suppose, slagged off fabric a bit in kind of a perfect way. yeah.

Lee Brownhill (41:40)
Yes.

I've got a feeling I know exactly which one you're on about. Yeah.

Yeah, ⁓ there's two very distinct parties

I see on certainly LinkedIn ⁓ around Fabric at the moment. So it's interesting and yeah, I'm keen to see how it all pans out with the whole Fabric tool.

David Morgan-Gumm (42:03)
Yeah. Yeah.

I mean, it's an exciting tool. ⁓ know, what I've seen that it can do is amazing, kind of bringing everything into one solution. ⁓ But like you said, you know, I don't know if it's daily at the moment.

Lee Brownhill (42:23)
Mm-hmm. Mm-hmm.

David Morgan-Gumm (42:32)
But you know, when I looked at it about a year ago now, I it was actually last September, looked at it in a bit more detail. It was, you know, changes were daily. There was just too much going on to be able to, I suppose, the moving our production systems to the business into that product. There's a balance, we've already mentioned the balance, but definitely a balance between

Lee Brownhill (42:52)
Yeah.

David Morgan-Gumm (43:02)
risk and reward with these new technologies.

Lee Brownhill (43:03)
There is, yeah, for sure. And that's precisely, you know, that's the situation we were in at the time with the managed instance. We were partnering with Microsoft and they were keen to get us on their platform, their new Shiny platform as they were at the time, showing my age again. But they were new Shiny and they were almost like for like compatibility. I'm gonna use some quotes there for obvious reasons, but they were as close, you know, as we had at the time.

And

there were things, like you said, like you said, with fabric, that things can change continually. And that's difficult to manage in a professional environment where things have to be signed off and risk assessed and change controlled and all that thing, the sort of thing. So it's a really hard sell to get the shiny new stuff sometimes.

David Morgan-Gumm (43:47)
I

agree. And I suppose that takes us on to ⁓ one of our last talking points of the future of data. ⁓

Lee Brownhill (44:00)
Yeah.

David Morgan-Gumm (44:01)
Fabric can be one of them, but are there any kind of trends going on at the moment that you're really excited about? Oh yeah, yeah,

Lee Brownhill (44:08)
Yeah, I mean, obviously my eyes were always going to be around the data platform, typically

databases in general. So the one thing I'm fascinated about is the rising popularity in Postgres at the moment. The rise is something actually quite astronomical and special. I'm really interested to see just how far it goes because more and more people are switching or certainly new.

David Morgan-Gumm (44:25)
Okay.

Lee Brownhill (44:35)
products, new environments, they're spinning up. They are looking towards database technology such as Postgres now. And the one thing that stuck me going back to Red Gate Summit again, I watched a session by Grant Fritchie talking about Postgres and it really piqued my interest in some of the things the platform can do. And I'm a complete novice obviously with that platform because I've never had any real world sort of experience in a professional capacity.

David Morgan-Gumm (45:00)
Mm-hmm.

I'm in the same boat.

Lee Brownhill (45:06)
But I'm keen to change that in the future because it's really interesting to me and some of the things that from a high level I've heard about it can do, you know, I'm really interested and it sounds really exciting. So I'm just really keen to see where that whole trend goes and I don't even think you can call it a trend now just because it's had such a rise for such a long time, so a pain rise.

David Morgan-Gumm (45:24)
Yeah, it has been around for a while.

Lee Brownhill (45:32)
Yeah, but it's really starting to become a talking point in the data world, which is always cool, right? I always like new things and new ways of doing things and new products and shiny new things basically. So yeah, I'm really interested

in that.

David Morgan-Gumm (45:50)
One of the things I've been seeing a lot more about recently in the same boat is ⁓ BigQuery. It's BigQuery, isn't it? The Google one. ⁓ I've seen a lot of adoption for the Google Cloud Platform in the last couple of years, especially within smaller companies.

Lee Brownhill (45:52)
Thank

Yeah. Yeah.

David Morgan-Gumm (46:12)
know, that all startup, you know, startup environments ⁓ building into the Google ecosystem. not really sure where, you know, whereas it all fits in, you know, in the future. Cause I think you've got AWS, which I think is used.

I think interchangeably with Azure with much larger companies. yeah, the take up on the Google Cloud Platform has been quite stark to the same sort of rise that you mentioned with Postgres. And the services that Google are offering on their Cloud Platform now associated with data processes and development and architectures, it's coming a long way. And they seem to be quite far at the forefront of new technology.

Lee Brownhill (46:34)
Yeah.

Yeah.

That's what good.

David Morgan-Gumm (47:01)
in that regard. And then especially with AI as well, you know, I, I, I use Gemini for everything. I'm not using Claude, I'm not using ChatGPT or, you know, any of those. Gemini for me is, has just outperformed all of the rest of them. I've been playing around with, um, Google Cloud myself. Uh, I said for, you know, for a couple of months now, um, just

Lee Brownhill (47:07)
Okay.

Right.

David Morgan-Gumm (47:26)
you know, with SQL squared and the hosting of certain products and processes. And the way Gemini is built into that system as well. know Copilot is built into Azure and they've been doing a lot with that recently. But I just feel like the modern feel ⁓ and that, I suppose that innovation that Google are pushing forward within the data space on their platforms has been quite impressive over the,

Lee Brownhill (47:37)
Yeah.

David Morgan-Gumm (47:55)
recent years.

Lee Brownhill (47:56)
Yeah, it's interesting because I'd say the Google side of the cloud world is probably the area. Well, not probably. It's certainly the area I've had the least exposure to in terms of coming across it in various roles. So Azure is my safe, happy place. I'm never going to deny that. I know Azure well, very well now, and I feel very at ease and comfort in it. But AWS is a

David Morgan-Gumm (48:03)
AWS

Lee Brownhill (48:21)
obviously one of the biggest ones out there now, I believe, in terms of uptake and

David Morgan-Gumm (48:24)
is huge.

Lee Brownhill (48:26)
just sheer numbers. And I think it's quite clear to see that when I attended the ADS Summit not long ago in London, ⁓ it was a whole extra level of ⁓ investment available to that business, right? And just the sheer size scale and the amount of things that were going on in that event was something quite...

David Morgan-Gumm (48:27)
Yeah.

Lee Brownhill (48:48)
special to see in real life. It's kind of crazy.

David Morgan-Gumm (48:50)
Yeah. Yeah, it's an area where I feel like Azure is falling behind.

a little bit. I feel in the community and things that I'm kind of seeing online, just seeing much more innovation within the AWS and Google space rather than the Microsoft space. Although, you know, Microsoft build was the other week and the conference that they had, I watched that and that was very, very promising. You know, lots and lots of new features, especially coming to

Lee Brownhill (49:15)
Yes.

Yeah.

David Morgan-Gumm (49:24)
the office suite and to help businesses in general. Whereas Google I.O. very, very impressive, much more focused on like the consumer, you know, rather than developers rather than businesses. So, yeah, the different platforms and companies are clearly taking different approaches. ⁓ Yes.

Lee Brownhill (49:34)
Okay, interesting.

Yeah.

Yeah, but I think these things are cyclical, right? I mean, I know Microsoft's having quite a transformation of a transformative time at the moment in terms of ⁓

how it's approaching various things and where its focus appears to be. And I think these things will just go in cycles and eventually course correct. And, you know, that's just how these things seem to work in my experience. Yeah, exactly.

David Morgan-Gumm (50:01)
Yeah. Need to wait for the next trend.

Yeah. I do think, you know, I'm very, a lot of people, especially developers, ⁓ can be...

You get opposite ends of the seesaw, I suppose. Some are very, very excited by AI and what it can provide and do, and I'm on that side. But others can be very, very skeptical. I have got a couple of people in my team that are a bit worried about it, to be honest, and resistant to pick it up. Because... Yeah, yeah, yeah.

Lee Brownhill (50:31)
Yeah, same.

And I think it's natural. I think it's natural because

you only have to open any social media around the tech community at the moment. And it's the number one thing everyone's talking about. And I think the key thing is to embrace it ultimately. And I think certain roles are obviously going to evolve and change with the advent of AI. It's just going to happen. But one thing...

David Morgan-Gumm (50:55)
⁓ yeah. Yeah. Yeah.

Lee Brownhill (51:15)
that I'm keen on focusing on is ⁓ abstracting myself out of the code almost because I used to consider myself quite a lover of coding, but those days are slowly going because they were always outperform most of us in terms of being able to just generate code quickly. You'll always obviously need that expert eye over the code to check everything is right because...

David Morgan-Gumm (51:20)
Yeah, definitely.

Lee Brownhill (51:43)
Because the one thing that AI cannot do is take in and understand business requirements and translate that into the tech. I think that's where our skills need to flourish really now in terms of the actual bare bones tech. I think it's about interpreting business goals, business needs into the tech. I think that's where our skills need to evolve into more.

David Morgan-Gumm (51:55)
Mm-hmm.

Yeah, no. I totally agree. Totally agree. ⁓

Lee Brownhill (52:09)
But yeah, it's a worrying time for a lot of people, I believe, at the moment. I read a lot of, you know, everything's doomed and things like that. It's

natural, it's natural, right?

David Morgan-Gumm (52:21)
It's a difficult thing to understand as well. Like I'm not going to pretend to understand how it works in the background. But you you hear things of like, I can't remember who it was. Some US audit process that went on about over an anthropic and open AI a few weeks ago.

Lee Brownhill (52:46)
Okay.

David Morgan-Gumm (52:48)
And the output of it was literally the people that working on it don't even understand how it works. ⁓ so that, you know, words like that become, you know, a little bit worrying. ⁓ But people, I think, are afraid to adopt stuff that they don't necessarily understand. And I think this whole AI thing ⁓ is probably the most complex new technology that has

Lee Brownhill (52:52)
Okay. ⁓

Yes.

David Morgan-Gumm (53:18)
probably ever come out and hit the world by Yeah, security is a big thing. Yeah, yeah. I mentioned the model context protocols earlier and linking it to SQL Server. If, you know, the audience people out there, people listening to this, you start working with these technologies,

Lee Brownhill (53:21)
I agree. And the security ramifications are obviously understandably and rightly so a hot topic around AI as well.

David Morgan-Gumm (53:48)
you've got to your own ⁓ AI subscription self-hosted in your tenant of using as your... It might not seem worth as a for a business to pay for copilot when everyone just gets it anyway, but you can't link these openly available AI endpoints to...

your systems using MCPs. mean, you can, technically you can, but you shouldn't because it gives you a window into your data warehouse and your processes. But if you have your own, I suppose, self-hosted tenant and you pay for that yourself in the cloud, then that window opens to you. it was the reason I've started working with the MCP stuff and doing proof of concepts on it.

Lee Brownhill (54:19)
No. ⁓

David Morgan-Gumm (54:43)
because I think the context that you can provide these things is, it's just, I've found to be insanely powerful, much, more powerful than I thought it was going to be to begin with. ⁓ But the security aspect is a huge talking point. know, the business that I work for at the moment, they looked at purchasing Co-Pilot

And they decided not to do it. It didn't provide them enough value for the cost at the time. But when you've got these additional technologies and these additional layers on top, it then reaches a new height.

But a lot of people, because it's a very, very new technology, a lot of people don't know about them, don't know about these, the new things coming out, because it's coming out at such a fast rate. We've talked about change on a daily basis with your manage instance and things like that. This is definitely that. But yeah. ⁓

Lee Brownhill (55:37)
Yes, for sure.

David Morgan-Gumm (55:42)
I think that that's all of our talking points that we've gone through. thank you so much, for sharing your insights.

Lee Brownhill (55:48)
Thank you, David.

David Morgan-Gumm (55:51)
I hope your goals of getting much more involved with the community go really well this year. It's been ⁓ great chatting with you. I look forward to seeing you finding Data Friends podcast episode. I'm sure it'll be a very interesting conversation to listen to. ⁓ Good.

Lee Brownhill (56:04)
Yeah.

Thank you very much.

David Morgan-Gumm (56:10)
And thank you all for tuning in. Please, if you enjoyed the episode, like the video, subscribe to the channel to hear more about the world of data and AI. And I'll see you next month with episode five.

Cheers Lee, thank you.

Lee Brownhill (56:29)
Thank you.