David Morgan-Gumm (00:00)
Hey everyone, and welcome back to SQL Squared, where we chat about all things data development and the cool tech shape in our world. We love bringing on awesome folks to share their stories and insights and today is no exception. Today, I'm thrilled to be joined by James Reeves. James's career is a masterclass in the evolution of data roles. He's been a SQL Server Consultant, a Data Warehouse Architect, a Data Engineer and Analytics Engineer, and even became a Director of Analytics. This incredible breadth of experience gives him a unique perspective in the industry.

which he now applies in his role as an AI engineer at Brabner's. Beyond his technical chops, he's a huge voice in the data community and is a passionate advocate for building inclusive tech spaces. James, it's great to have you on the show.

James Reeves (00:43)
Thank you for having me. We've been chatting for a while, so it's good to get on your podcast.

David Morgan-Gumm (00:47)
Yeah, we finally got the recording in place. This is going to be an insightful chat. We'll be diving into the current state of the data engineering platforms, discussing the ever present challenge of cloud cost efficiency, which loads of people will be familiar with, and looking ahead at what the next decade of tech careers might look like. We're also having a really important conversation about neurodiversity and equality in our industry.

So plenty to get through. Very first topic, we're looking or chatting about the great data platform debate, Databricks versus Snowflake. Myself, I don't have much experience with Databricks or Snowflake, but I know that you've had a bit of a play around with Snowflake and you're a bit more experienced in the Databricks space. So.

You know, there's always a huge debate, a huge conversation about which platform you should go to. And obviously Fabric is coming into the fray now. From your perspective as an engineer, kind of what makes one shine over the other or kind of what are the pros and cons in your experience?

James Reeves (01:52)
Yeah, I love Databricks in that it's whichever language you're competent in, it's SQL or Scala or any of the many which are available. The performance is equivalent, which I think is brilliant. You don't lose anything by writing in one or the other, especially when it comes to things like pivoting data, SQL.

David Morgan-Gumm (02:09)
process.

James Reeves (02:13)
is not good at that. It's never been very good at that. You should always do it in the visual layer usually, but sometimes you do need to do it in the data layer. And that's when Python just, it's just much better at that. So the fact that you've got all those and you can run a notebook top to bottom and just swap between the languages is amazing. So that's one of reasons that I would always

have a strong positive vote for Databricks. In terms of Snowflake, I have not been an administrator of, but I've been a user of. So I've not had a chance to set up the, to have a play setting things up. But what I have experienced from using Snowflake is it's got some very cool, again, around sequel, it's got some optimizations which are in Databricks now, but Snowflake had them first.

which just make writing SQL so much easier. you know SQL, it's things like if I do an alias, a column on line one, I can then reference that alias on line two. So if I was to add this column times that column is known as this, I could then do this minus five.

David Morgan-Gumm (03:12)
Heh

James Reeves (03:36)
and it worked, it knows it. Yeah, it's super, it just cleans it up so much. But then in terms of the, and again, they're both developing, the visual analytics side, that's where I would always, historically have been, Snowflake is great for that. For self-service analytics, it's just a really nice layer for users to use.

David Morgan-Gumm (03:37)
I can imagine that being really helpful.

James Reeves (04:00)
than Databricks is now. The thing is, it's always in competition. So Databricks, Snowflake, not to forget the, it's known as legacy tech, which makes people think it's old and outdated and going out, but it really isn't. So things like Azure Data Factory, SQL Server, and Power BI, there's a classic stack and it's very powerful.

I actually do a talk showing people how to spin up that classic stack in virtually no time at all. It's about, it's sub 10 minutes to deploy and then maybe an hour or two to configure and then you could literally be proof of concept in under two hours, which is I think quite a cool thing because then you can spin up, try something, if it doesn't work, spin it down.

Um, it's not months and months of commitment. then if you find that it doesn't work, then there's a lot more lost. Whereas if you can get it up in, let's say a day, you two hours, I've done it a lot of times, number of times, but easily do it in doing a day. Um, so yeah, great data platform debate. They all have, um, positives. I would say one thing I would say is, um, sometimes it's, um,

David Morgan-Gumm (04:56)
Mm-hmm.

Hahaha.

They do, they all have positives.

James Reeves (05:22)
look at the technology, look at the cost of ownership. When I'm working with, well, when I was working with clients, I took into account what their skill set was. So if they're a Python house or have used languages like that, then Databricks is almost certainly the answer. If they're a SQL house, then try and keep everything, you know, even to the extent of having a metadata driven framework so that

David Morgan-Gumm (05:32)
Mm-hmm.

James Reeves (05:49)
you can choose what gets downloaded and processed every night or every day or whatever purely using SQL. And that's best thing then because you can go, here's, yes it's an entirely new platform that you didn't have before, but it's based on a language you know and your team knows. But yeah, so very strong advocate of SQL.

David Morgan-Gumm (06:10)
So

yeah, know, the same. mean, that classic stack that you mentioned is something that I still use today on a daily basis. The company that I work for, we've got flows and data storage and analytics all running within that classic stack. We're only recently moving to Fabric, which is still in its early days, even though it's been around for a few years. And we're finding challenges with it, but it's

is clearly built off the idea of Databricks, kind of with that Python element to it and the notebooks. But the, you mentioned that there's features that were in Snowflake that have come over to Databricks and there's probably been a few vice versa as well. Do you think these sorts of platforms will eventually become indistinguishable from each other or do you think they'll serve fundamental different use cases? Or is the use case the same but

the way that it's implemented is different.

James Reeves (07:06)
Silence. Why is there great debate? I don't know. Could they all become less, I think it might become less obvious which system you're using, Databricks or Snowflake or Classic or Fabric or anything like that. Just with the, how AI is getting to be just much more the first place you go rather than.

David Morgan-Gumm (07:07)
Hmm. ⁓

Mm-hmm.

Yeah.

James Reeves (07:29)
rather than it being an assistant to the side or a separate part, it's very much becoming the first thing. I want to look at this data in such a way and it's an AI query rather than going to Databricks and writing a query or going to Snowflake or putting in Power BI. So I think the the stacks might become more transparent.

I think the individual ones will still exist because they'll have specialties. So I think in terms of AI, machine learning, modeling, things like that, I think Databricks will probably be the front runner. And this is possibly my lack of experience in Snowflake. I don't think Snowflake has that kind of power, but it could be wrong.

David Morgan-Gumm (07:58)
Mm-hmm.

From what I've seen in tech

news, know, Databricks is quite at the forefront of kind of that AI and ML integration within a data platform, especially for analytical purposes.

James Reeves (08:26)
Exactly. And the other thing is you're locked in that you could use Databricks for a certain piece and then go to the Azure Foundry for your AI requirements and you can pick and choose. I like this bit, I like that bit. And again, could be my lack of experience. I think Snowflake is less like that. I think it's a self-contained universe.

David Morgan-Gumm (08:42)
Yeah.

Okay, great. mean, yeah, there was one last question that I was gonna ask, I think you've already answered, which is about kind of, know, what questions should people ask themselves when they're choosing a data platform? Because I think I've spoken to quite a lot of people in the industry throughout the year. People are getting very excited about AI and the AI implementations that can be added.

to these sorts of data platforms. And people are thinking about switching from that, for example, that classic data stack and taking on larger projects to change the technology of the data platform they use. I mean, I think there's a few crucial questions that people need to ask themselves when deciding on a platform. I think you've already covered a couple of them. One is the tech stack that you wanna work in.

and the other being the degree of AI implementation or potential future implementation within that platform. Are there any others that you think people need to consider?

James Reeves (09:53)
⁓

cost of ownership and maintenance. Do you want, I think what you might see as well coming in the future is kind of the analytics as a service you might get. ⁓ So for a while now you've had Power BI kind of as a service where an external company will build your Power BI reports for you, et cetera, et cetera. And we've always had the external companies will come in and

David Morgan-Gumm (10:07)
Mm-hmm.

James Reeves (10:21)
build the data platform for you, I think we'll come to a point where it'll be bring your data to our platform. the go-to, the consultancy or the company would have a data platform, they pay for the infrastructure. The company needing a data platform would just make use of somebody else's.

I that I've never come across that, but I think that might be a thing that comes. And one of the reasons for that is the uncertainty of cost of data platforms. How much is it going to to spin up? How much does cost to keep running? What features does it have? How do I do the licensing? It's actually quite a, there's a phrase I heard many, years ago, which was, if you're not a software company, don't write software.

David Morgan-Gumm (10:57)
Mm.

James Reeves (11:11)
That was more around end user computing and when Visual Basic was the king of making processes a bit faster. And if you think about the requirements to build and run a data platform stack, you could apply the same logic, is if you're a company that does, if you're a call center or if you're a tool shop or if you're a,

David Morgan-Gumm (11:17)
You

James Reeves (11:36)
factory. Do you want to be in the business of having a data platform? And, know, historically, yes, they do. They have a department that does it. I've got in my mind that I could see that changing, that you would

David Morgan-Gumm (11:49)
yeah,

they can kind of buy that service from other people. I see where you're coming from now. Yeah.

James Reeves (11:53)
Yeah,

that's what I think we'll go to. So then the query of just starting a modern data journey, that might change that question. Instead of do we wanna build or who do we want to build or what it's, right, which one do we buy? And then it's a purchasing question rather than tech building question.

David Morgan-Gumm (12:13)
That's a really good way thinking about it. I hadn't really considered that.

James Reeves (12:17)
It might not come, it might be because they're so, that's the weird thing. It's always been regarded that a tech stack or what you need out of something is so individual that you have to build your own. But then if you look at Dataverse or the Common Data, what was it called before? We've got the many iterations of the naming of that, Common Data Service and the Common, but if you look at that,

people will always need account data, they'll always need people data, they'll always need invoices. Yes, each company's individual, but really they kind of have similar needs. Therefore you could buy something off the shelf and then just customize. So I don't know why that hasn't happened yet. So a bit off topic, but.

David Morgan-Gumm (12:45)
Mm-hmm.

Mm-hmm.

No, I mean, it kind of takes us perfectly onto the kind of the next phase of, you know, what one of the things that we talk about on this podcast quite a lot is cloud cost efficiency and the difference between buying and building your own and what people have to consider. So on the back of that, what are some of the biggest mistakes you see companies make that cause their ⁓

Cloud builds to go through the roof.

James Reeves (13:32)
Leaving defaults, deploy with defaults is definitely one. I can personally attest because I did that. I spanned up a SQL server, Azure SQL server for a user group demo, left it on the default setting, and then had a rather large bill that I was completely not expecting because I'd left it on, the initial one is something like 1500 pound a month.

David Morgan-Gumm (13:33)
You

Mm-hmm.

Yeah.

James Reeves (13:58)
And

the vast majority of clients I've worked with have not needed that level of power. So yeah, first thing is turn it down, turn it down to the minimum and then work up and can scale up, scale down as and when needed.

David Morgan-Gumm (14:12)
Work up from there.

James Reeves (14:16)
Just repeat the question again. what's the, what's the.

David Morgan-Gumm (14:18)
Yeah,

so the biggest mistakes you see companies make when they're setting up infrastructure on the cloud that causes their bills to go through the roof.

James Reeves (14:29)
So one is defaults, suppose you could say it either way. The other one is, especially when comes to SQL Server, when you are making things fast on TIN, on your machine that you have in your office, you don't really care if the CPU is maxed, if the network controller is maxed, it doesn't really matter. What you care about is when a query is run,

how long does it take to come back? That's really the, and if you're tuning the query, yes, you will look at IO and CPU and things like that, but for the purpose of tuning, no other reason. And then when you go to cloud, everything you've tuned for the machine in your office, it's not necessarily the right way to tune when you go up in cloud because

you burn CPU, that's gonna cost you. If you have a lot of data moving around, that's gonna cost you. and because I've been in the industry for a while, I don't know, I'd like to go, I'd like to see what it's like for someone getting into SQL tuning now, even on, on Pram. And I think the...

The focus being on efficiency, not on delivery speed. So if something takes a millisecond to come back, but uses a load of CPU, or something takes two milliseconds to come back and doesn't, I hope nowadays the choice would be, well, do know what, difference between one and two milliseconds is, it's nothing, a user's not gonna notice that difference, but it's a lot more efficient, so let's go with the.

the two millisecond version. Whereas when I was, when on-prem was my main daily thing, it was just how fast can we get it back? I didn't really care about CPU because unless you're honest on a box which is constrained because there's lots of users, efficiency is just not a thought process.

David Morgan-Gumm (16:27)
Yeah.

James Reeves (16:41)
I I've attended a user group session where they were looking at environmental cost and that was a consideration at that point. But it's the same thing. If you're more efficient for the planet, then you're also probably more efficient for the cloud in terms of costs. So that's a win-win.

David Morgan-Gumm (17:01)
Yeah, that's very good point.

Yeah, they talk about, you know, why is the word leaving my head? The kind of like ecological impact of AI and the massive data centers that are being built just to, you know, return you a reply to your entry to chat GPT. I know that, a huge talking point at the moment in the news.

about the impact that's having. But that's just a much bigger scale to exactly what you're talking about there.

James Reeves (17:28)
Yeah. I think I saw in, might have dreamt it. Not that I dream of visual portal. I'm sure when I looked at SQL server, one of the stats was an eco rating. Pretty sure I saw that come in, which was quite interesting. Yeah.

David Morgan-Gumm (17:35)
Hahaha.

Really? Yeah,

they're probably doing it with the AI stuff. And then I know that, you know, one of the things I wanted to chat about with you, because we're, you know, we're both SQL Server experts, you know, and gonna gone over what the larger, more modern data platforms are, you know, does SQL Server still play a major role in, you know, modern architectures, or is it becoming a bit of a legacy system? Because like, like you've

You know, we've just mentioned about AI and everything. See what server 2025 is coming out soon. It's rumored and it's been hyped up quite a lot to have some decent AI features. I'm not really sure what they are, but I'm assuming that's why that, you know, ecological score has been added.

James Reeves (18:25)
Yeah, potentially. Yes, was Azure SQL Server I was looking at, which is not on. It's always weird, isn't it? It's the latest edition, but it's not necessarily, it's not actually.

David Morgan-Gumm (18:35)
No, yeah, it's a bit behind the like the Azure SQL databases. Definitely. They're quite limited, aren't they? I use a managed instance at work. Found it to be so useful. Yeah, just like you said, you know, the ability for it to just constantly stay up to date with all the newest features. You don't have to do all the management of the VM and everything behind the scenes. It just makes a huge difference and it gives you all of pretty much, well, yeah.

James Reeves (18:41)
Yeah, and again. Okay.

David Morgan-Gumm (19:01)
Pretty much all, I'd say 95 to 98 % of all of the features that you'd get from a kind of a classic SQL server install on a VM. I'd recommend Managed Instance to everyone. I think I mentioned it on every episode. ⁓ But yeah.

James Reeves (19:15)
I would recommend

Azure SQL DB and not Managed Instance. I don't know why, I don't really use Managed Instance, it's because I don't need the latest, greatest features, I need the core set and the fact that the basic setting is five pound a month. That's a fixed cost.

David Morgan-Gumm (19:19)
No, fair enough.

Yeah. Yeah, it is.

James Reeves (19:40)
It's great that you can't go over that cost. If you set it at that power level, it will cost you five pounds. So having that certainty can take a lot of fear out of going to the cloud. It's not always gonna be a thousand pound bill this month. No, if you configure it as a basic, it will be five pound. Yeah, and that, well actually.

David Morgan-Gumm (19:47)
Mm-hmm.

Yeah, you're guaranteed, though, really.

James Reeves (20:03)
the, what is it, is it eight databases now per person, per resource group or per subscript?

David Morgan-Gumm (20:09)
They can't talk

to each other though, can they?

James Reeves (20:11)
No, and then that becomes a change of engineering pattern that you separate things with schemas rather than databases. So they're all on the same database or uses your data factory to talk between two, but then you pay in to move data around. Yeah, so yeah. So when I was on premise, I did a lot with link servers.

David Morgan-Gumm (20:24)
Mm-hmm. There are the challenges that I came across.

James Reeves (20:32)
yes you're right that is that does become more problematic.

David Morgan-Gumm (20:36)
There was

quite a few issues we had with networking with the Azure SQL databases. I think they, I don't know if they are anymore, but when we started working with them three years ago, I think there was quite a few operations, especially to other resources that could only be run over a public endpoint. And the business was adamant that it had to be a private endpoint and running through,

a setup virtual network within Azure. And so we ran into a lot of problems there, we're trying it, but we looked at managed instance and managed instance supported the private endpoint setups.

James Reeves (21:12)
Yeah, keep it all in a long time and yeah.

David Morgan-Gumm (21:14)
Yeah.

So yeah. SQL Server in general, what do you think? Do think it's going to be around for a lot longer? mean, SQL, the language will be. But SQL Server as a product, I mean, even with Fabric, you think maybe we're going to talk about the next 10 years in a bit? Do you think it will be phased out in the next 10 years?

James Reeves (21:34)
difficult and people have been predicting the death of it for a while. It never happened. I don't know. think...

David Morgan-Gumm (21:37)
Yeah, they have, yeah.

James Reeves (21:41)
Seeing how much Microsoft's doing with Postgres as well. I never know if Postgres SQL or Postgres , Postgres , yeah, it's a weird one. Yeah, can I see the death of SQL Server? No, particularly because there's just so much out there which is running on it. Do I see it morphing and becoming something else? Potentially. I think I have to.

David Morgan-Gumm (21:44)
Yeah, yeah, yeah, yeah. Post-gres, I always say.

That's a good point. 2025

might be the start of that. We don't know what these AI implementations are gonna be like.

James Reeves (22:08)
Yeah.

Exactly.

quite a few, over the last few years, there's been things bolted onto SQL that don't feel native. They feel like, oh, we need to have this functionality, so we'll just stick this on the side. And it's kind of like, oh, it doesn't feel like it's part of the product. things like Databricks and Snowflake, have these abilities. So you want SQL Server to have these abilities. No, I think if you're in the SQL Server world and you need to be able to do a certain thing,

then you'd probably be perfectly happy spinning up a resource in Azure that handles that use case rather than having SQL Server natively build it in.

David Morgan-Gumm (22:45)
Mm.

Mm-hmm.

Yeah, that's a really good point. I agree. I think it will evolve rather than go away. Yeah. ⁓

James Reeves (22:54)
Yeah, but would it be post-graph or what? I

don't see it being able to run things like Python and all these other engines. That's such a difference. I think it will still be a sequel engine. It just may not be the same engine we have now, which is awesome at the moment. And I don't know. I think it...

David Morgan-Gumm (23:08)
Yeah, it is.

James Reeves (23:22)
I think we're maybe in a diminishing returns time where it's difficult to make it any much better because it must have a lot of legacy built in.

David Morgan-Gumm (23:32)
Yeah, I think there'll be a lot of limitations compared to the other more modern platforms, especially when they start to evolve a lot more to the modern needs.

James Reeves (23:42)
In

terms of SQL language I mentioned on Snowflake, Databricks, they're able to iterate the recursive CTEs, I think just came into Databricks recently, but there's no new SQL server, no Databricks version. That infrastructure of Spark Engine and being able to...

David Morgan-Gumm (23:45)
Yeah.

James Reeves (24:06)
implement new things just by changing the engine and then your end users can pick and choose which one do I want to use. It just enables development and improvements far faster and they could probably do a really big breaking change between versions and it's up to the user if you want to move to this new spark engine that does this brand new thing that everyone loves.

David Morgan-Gumm (24:14)
Mm-hmm.

James Reeves (24:33)
you're just gonna have to spin up another instance and you can try it. I think that's easier. Yeah, so if you have a new version of SQL Server that you need to check for compatibility, doesn't feel like on Databricks, you could just have a notebook and run it on that engine, run it on that engine, check the output really easy. SQL Server's a lot more involved to be able to do that.

David Morgan-Gumm (24:38)
Yeah, flexibility is huge win.

James Reeves (24:59)
So whether SQL Server will move towards that kind of, I mean, it's kind of got that, you could argue it's kind of got that with the, you choose your compatibility level, but whilst it's similar, I don't think it's the same.

So long, long answer to, think it will still be around. I don't think it will look the same as it does now.

David Morgan-Gumm (25:13)
You're right, you got a very good point there.

Great. Yeah.

All right. This is a good point to pause for a second, I think. So just message to the audience out there. If you're enjoying this discussion on cloud costs and data platforms, whether it be Databricks, Snowflake or SQL Server, make sure you're subscribed and use those notifications so you don't miss future episodes of the podcast. And if you have a question that you'd like us to tackle on the podcast,

Send it over to mailbag@sqlsquared.co.uk.

Okay. So onto a bit of a fun section. We're going to chat about meetup mishaps and conference confessions is what I've ⁓ called this section. I know that you like to be involved in the data community. You've got your own meetup. I you've been to and spoke at meetups before, I'm sure you've been to many conferences before.

So want to get a bit of an insight into your experience at these events. So they can't always go smoothly, surely. Is there a memorable mishap that you've experienced?

James Reeves (26:18)
Alright.

I'm trying to think of that. I just want to point out as well, geeking out on SQL is fun. We're not moving on to fun. That was fun before. I can always geek out on SQL. So things that have gone wrong at conferences. I did attend a SQL bits in Manchester a few years ago and it wasn't really 2018, 2025 now, seven years ago, possibly.

David Morgan-Gumm (26:32)
Ha ha ha ha ha.

is that the 2018 one.

know, it was in Manchester in 2018. I ⁓ didn't go to the ones just after COVID, so I'm not sure.

James Reeves (26:51)
That's all.

Yeah, I mean it's been there a couple of times but I'm not sure if it was that long. It feels about right. And because it's such a big conference and there's so many speakers, so many attendees, and a great conference by the way, I recommend if you are new to the world or long in the tooth, if you've never been to SQLBits you should. You'll learn so much. It's a funny name as well, SQLBits, it's not, sequel is not the focus.

David Morgan-Gumm (27:16)
Definitely.

Not anymore, anyway.

James Reeves (27:23)
No,

not anymore. But the name's got such a like a brand behind it that you couldn't really change it. But yeah, I hope it doesn't put people off because they talk about data in all flavors. Which I get so much from it. The technical mishap was they because they have so many speakers in how many places in the country can you get 12 speakers?

at the same time with between 50 and 100 attendees for each session and you need to have them together. Can you imagine the complexity of that? So SQL bits they have big heavy black divider curtains and they work really well for sound deadening but there's a little bit of leakage that goes up and over because there's no lid.

because in Manchester they tried having a lid, put a big curtain over the top and it just became an oven. was, you hit all those people and you're completely surrounded by this really heavy fabric. It was so hot. I mean, that's, if that's the worst mishap, then that's not too bad, is it? My own personal mishap.

David Morgan-Gumm (28:29)
Yes, it's black as well, isn't it?

No, that's not too bad at all. Not too bad at all.

James Reeves (28:42)
was fun, I was giving a ⁓ presentation to a user group, I can't remember who or what now, but whoever set up the Teams call hadn't locked it down, so everybody could turn the video on, everyone could turn, everyone could share, everyone could use a microphone, et cetera, and somebody, a young somebody thought it was hilarious to take my feed.

put like a half second delay on it and feed it back to the session. And I don't know if you've ever experienced that, the cognitive dissonance of hearing yourself, the half second delay. We ended up having to abandon the session because we couldn't take the permissions off or we didn't know, they didn't know at the time how to restrict people being able to do that. So yeah, I tried, I battled on for like five minutes.

And it got to the point of like, can't, I can't think because I'm just hearing me. Um, so yeah, that was interesting, but, um, it was funny. we, um, the, we just reorganized and did it again. Um, and that's, that's something I want to, uh, I would say to, um, if you're thinking about speaking or, um, if you're just worried about a lot of stuff in your, in your tech career, just.

David Morgan-Gumm (29:37)
yeah, yeah. That is horrible.

Okay.

James Reeves (30:02)
it will come right if you can relax generally there's nothing nothing going wrong that you can't fix so

David Morgan-Gumm (30:08)
No, I agree. I agree.

And I found that with this podcast. Definitely. Just relax. You're Yeah, I know. Yeah. Just on SQL bits, just something I wanted to mention, both for you and the audience, recently recorded a podcast, which isn't yet released and editing at the moment as time of this recording.

James Reeves (30:12)
Yeah, yeah, just for likes of you, alright? You can cut bits out as well if you want.

David Morgan-Gumm (30:33)
with Simon Sabin, one of the founders of SQLBits. They have kindly given us a discount code to use at next year's SQLBits. for people that you may listen to the actual SQLBits Q &A, that will have come out by the time this episode is released. But if you go to SQLBits.com to get your tickets,

to the conference in Newport in April, I think it is, 2026, you can use the code sqlsquared2026 to get a small discount on your tickets. And feel free to use that as well, James, if you want it.

James Reeves (31:03)
Yep.

I volunteer at SQLBits. I'm one of the orange shirts. If you attend and see someone in an orange shirt with Sequelbits written down the side, we're there to help. Make sure the speakers have got water. Make sure they're okay and settled. But also make sure all the attendees know where to go. Yeah, I've volunteered for the last few years. I love it. It's like a holiday where I get to meet all my friends I haven't seen for the year.

David Morgan-Gumm (31:15)
dear.

Yeah.

Mm-hmm.

James Reeves (31:40)
And you meet up again at SQL bits. It's really nice.

David Morgan-Gumm (31:43)
Yeah, I'll have to make sure to look out for you in the next conference.

James Reeves (31:46)
I'm there, I've already my airbnb. The day they announced where it was going to be I booked my airbnb. That's my tip, if you're attending, book your accommodation as soon as you can. can be. But Newport's great location, was there a couple of years ago. What I am looking forward to is finding out what the theme for the party is. There's always a party on Earth. It is very good at keeping secrets.

David Morgan-Gumm (31:51)
⁓ nice.

Yeah, pretty quick. It gets expensive.

Yeah, Simon didn't give anything away to me on the yeah,

I mean, I the last time it was in Newport, it was Dungeons and Dragons themed because they've got the massive dragon outside the conference. So it kind of made sense. I think they're going for a similar theme this year. That's kind of what I got from him. So a bit of an inside scoop. ⁓

James Reeves (32:25)
Nice. Okay.

You hear so many people, they all know the answer and they're all different answers. Somehow they keep it secret really, really well.

David Morgan-Gumm (32:40)
No, I know

they do. It'd be close to the chest. One other question then. So if you had to give a talk on a non-technical subject, what would it be?

James Reeves (32:53)
mental health, the importance of looking after it. It's just this simple, everyone always has this idea that there's poor mental health but forgets the other side that there's good mental health. It's like sometimes you're ill, you have a flu and sometimes you're not. Physically, think mental health is the same, it's sometimes you're okay and sometimes you're not.

Yeah, to think about it holistically, I'd say that's important. Yeah, so I don't talk on mental health, particularly neurodiversity. I was diagnosed with ADHD, I can't remember when it was, recently, two or three years ago. And the more I've learned about it, the better I've been able to deal with some of the challenges it brings.

and the better I've been able to leverage some of the benefits it brings. Very much not a it's a superpower, it's not. It can be useful but it can also be horrific. So yeah like to give talks about just to give what I've learned and share it people because they can take some of it and make improvements if they are

David Morgan-Gumm (33:48)
Mm-hmm.

Hehehehehe

James Reeves (34:10)
struggling themselves or they know somebody who has neurodiversity and doesn't know how to talk to them or help them or something like that. yeah, that's my, as I've just spoken a lot about, clearly a passion. I like to talk about neurodiversity.

David Morgan-Gumm (34:27)
No, that's great. I mean, it's really good to have advocates for that sort of, for that space really within tech. You do get people talking at conferences about mental health and neurodiversity within the workplace. But I think so much focus is on the actual technology that you're implementing and not, there's not enough around the actual individuals within a team doing the thing.

and how they work together and how you get the most out of people as well. Just a follow-up question on that. Are there any kind of like unique strengths and viewpoints that you think neurodivergent people bring, especially to a tech team?

James Reeves (35:15)
Yes, I mean, it's kind of a weird, everybody's neurodivergent. everybody's neurodivergent. That's the point. If you take the term neurodivergent. So everybody is different. Everyone's an individual. So if you have a team of people, they will all be different neurologically. So they'll all be neurodivergent.

David Morgan-Gumm (35:18)
We like weird.

Mm.

James Reeves (35:36)
I know get some, obviously some people have a recognized or diagnosed as a weird term as well. They have a recognized neurodivergence, like a categorized neurodivergence. But yeah, everyone brings unique strengths. I would say some of the strengths of things like autism, ADHD is the ability to

David Morgan-Gumm (35:50)
Mm-hmm.

James Reeves (36:04)
focus, get on a topic and stay on that topic for prolonged periods, like hyper focus on it. But that can actually be one of the weird things for people who don't understand that kind of neurodivergence is ADHD is where you can't focus. Well, yeah, it is, but it's also where you hyper focus. It's not.

David Morgan-Gumm (36:11)
Mm-hmm.

James Reeves (36:26)
It's because we always look at the negative, you can't focus or something like that. It's more that you're focusing, the way you focus is different to somebody else. It's not all, if you think about a neurotypical person, sometimes you focus and sometimes you don't. think ADHD is exactly the same, it just can be through a magnifying lens.

David Morgan-Gumm (36:43)
Yeah, yeah.

Yeah, kind of the extremes as well of either end. Yeah.

James Reeves (36:50)
Yeah,

strength being you can you can really refocus negative is it has to be interesting. That's I heard it. I heard it phrase that the neurotypical brain is reward focused and neurodivergent, particularly ADHD. Because I'm not that familiar with some of the other categories is is interest based.

David Morgan-Gumm (36:58)
Mm-hmm.

James Reeves (37:18)
If I'm not interested, I'm not doing it. Which can be really challenging. Because sometimes you have to in a workplace. got to do stuff that's not interesting. yeah, so yeah, the unique strengths and viewpoints. And yeah, so actually the other thing to mention is people can, if you have a variety of people in a team, they all bring different perspectives on a problem. I had,

I've had a number of examples in my career where we need to do this and this and that, and just having the ability to go, or what, or so what. it's really bad if this happens, like, okay, why? To who? Why is it bad? And then the ability to reframe it and go, well, cause we don't want to do it. But you don't want to do it because it's bad and it's bad because you don't want to do it. That's that's a non, that's a non thing.

is it actually a problem? We've got this other problem which is costing us money and all making people unhappy. Should we focus on that rather than this thing which is it's bad but no one can articulate why it's bad. It's just it's something we shouldn't do. Having that alternative viewpoint sometimes and that's just one example. The point is having an alternate take on things can be a real strength.

David Morgan-Gumm (38:18)
Mm-hmm.

Yeah,

I agree. It can be very, very powerful. think there's a couple of tips for managers and leaders out there as well within the tech space.

James Reeves (38:38)
If I can give one tip for managers, the really strong one, because I think one of the worst parts of ADHD, there's something called rejection sensitive dysphoria, RSD. And the way I try to explain to people, if you ever drank too much alcohol, woke up in the morning, woke up in the morning and you have that dread feeling of who did I say this to or what did I say, what did I do kind of thing. RSD is that.

David Morgan-Gumm (38:54)
quite a few times.

James Reeves (39:07)
all day, every day, and it sucks. It's one of the worst parts of ADHD I say. it comes, and so being aware of that, things like if you are organising a trip out, and let's say it's a social trip, and you have someone with ADHD, and I think autistic people struggle with the same, if they say can I come?

David Morgan-Gumm (39:09)
And yeah, I can imagine.

James Reeves (39:31)
and you say, if you want, if you want is the same as saying, no, I don't want you to come. Whereas a neurotypical would hear, if you want to come, can come. Neurodivergence tend to hear on what I particularly want you to come. If I did, I'd have invited you. So try and rephrase that kind of thing of, people can get, the other one is, you got a minute on a text message? Have you got a minute for your manager? I'm going to get fired.

David Morgan-Gumm (39:36)
Mm.

James Reeves (39:57)
Even if I've done nothing wrong in the world, can I have a minute is the dread sentence. Just being a little bit clearer, have you got five minutes? I want to talk about something, give context. Because that constant, and it's also bad for cortisol, the stress level of people. Just being aware of that can help you not do that.

David Morgan-Gumm (40:04)
Okay.

James Reeves (40:21)
and that'll make you have a much nicer relationship.

David Morgan-Gumm (40:23)
Yeah, definitely. I mean, there's, there's a funny thing where, um, we've got a senior manager in, um, in the department where I work. And if he ever needs to have a team meeting, he'll just create a meeting saying team update and then no information. And so everybody's then speculating for the entire day about what's happening. Somebody leaving. I don't know. Yeah, I just, yeah.

James Reeves (40:42)
Terrifying. No, the company's closing. That's terrifying.

David Morgan-Gumm (40:51)
He always gets feedback about it, but never changes how he does it. But yeah, that's an extreme example of that.

well, just what I suppose. Yeah. And one last question on kind of the neurodiversity and equality part. Looking at the, I was looking at equality in tech more broadly. What do you think? Do think there's any big hurdles that we still need to overcome as an industry? So like truly make it?

level playing field for people. Because I know, especially in tech, software development, programming, you see neurodivergent people, know, people with autism doing really well. You know, the, the hyper focus, you know, the ability to spit out applications and code that, you know, we wouldn't necessarily think of, but do you think, do you think there's any major hurdles to, to people for it, for it to be a level playing field for everyone?

James Reeves (41:45)
I don't know because it's really hard to come from my perspective. White, male, the age I am, the background I've got. I will see things as I've experienced them. I like to think we're doing really well, particularly in gender equality. I think we do quite well in recognizing neurodivergence.

David Morgan-Gumm (41:57)
That's fair, that's a point.

James Reeves (42:07)
But again, I'm coming from like SQL bits. do, they have talks on mental health, SQL bit and neurodiversity and gender equality. So I'm, I'm in this bubble where I think we are really good at it. But then I'm aware, I saw something on LinkedIn recently where the experience when you're actually in an underrepresented or a not fairly represented group.

David Morgan-Gumm (42:12)
video.

James Reeves (42:32)
your perspective, their perspective is very different because you know they're experiencing inequality. the single biggest hurdle

I guess if the problem is that someone like me is not aware of the problem, that probably is the problem. So how do you make people more aware? I don't know. I don't tend to read stats of male, female, neurodivergent, non-neurodivergent, sexuality, all the different variables.

David Morgan-Gumm (42:47)
That's a really good answer.

going to talks

at things like SQL bits and making sure that you're there for the full experience and to actually understand what's going on in the industry as a whole and not just there for the tech. You know, is something I recommend to everyone.

James Reeves (43:10)
Yeah, absolutely.

David Morgan-Gumm (43:13)
I'm not one to go to all of the talks around, you know, the quality and new diversity parts of tech. I do go there for the tech, but I do make sure to go to at least one or two in every conference. Just to be able to, you keep myself grounded within, you know, the area as well. you know, I'm a manager.

make sure that I'm all of the tips and tricks that I need to help make my team better as a whole.

James Reeves (43:46)
I

think it's the biggest part is being aware that there are different ways of managing. So I think you're doing a great, you're already doing a great thing. You're aware that there are ways to, or different ways of managing, just being aware of that is a great start.

David Morgan-Gumm (44:08)
I think everybody needs to be at that point at a minimum. Just be aware.

James Reeves (44:10)
Yeah, and

I think we're a lot better just generationally as we go through time, I think we're just getting better and better. It's not a tech thing, I think that's a world thing. We're becoming more aware of all sorts of things. Even saying things 10 times in the last sentence, which could do better than that. Actually, I've just thought of a tip and it's an ADHD brain where it's going all over.

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

Yeah

James Reeves (44:36)
Don't force people to have their cameras on. There are people who just don't have, it's just not a camera day and that should be okay as well. Everybody wants to be on the show all the time. So I think we should normalise, yeah it's not a camera day. Fine, it's not a problem. I get that it can promote or does promote closer communication.

David Morgan-Gumm (44:45)
Yeah, definitely.

Mm-hmm.

James Reeves (45:00)
But you've got to be aware that some people don't want to. Maybe they'll never want to turn their camera on and that should be okay as well.

David Morgan-Gumm (45:05)
Yeah,

yeah, especially within larger groups of people, know, large forum updates and things like that. Yeah.

James Reeves (45:12)
Yeah, 80 people what's the point in having you camera on ⁓

David Morgan-Gumm (45:20)
Pivoting a little bit, last topic of the recording, the next 10 years career growth, what direction do people need to go down to get the most out of their career over the next 10 years? There's a ⁓ huge pace of change at the moment. was something mentioned in the previous episode with Simon Fraser of he likes to stick with sequel.

because he doesn't want, doesn't, you know, he's struggling to keep up with, you know, the major changes that are going on in the industry, especially with AI. So, you know, this podcast is,

directed at professionals in the industry, but also people wanting to get into the industry, people still in education. So for a young person starting their education now, wanting, being ambitious and wanting a successful tech career, what advice would you give them to make the most out of the next 10 years?

James Reeves (46:21)
Good question.

I think you've mentioned some of it in what you said in the question actually, which is it is such a broad industry and you could be doing visuals in Power BI or Tableau, this is not a Microsoft sponsored podcast. So you could do visualization, click pyramid.

David Morgan-Gumm (46:29)
Mm-hmm.

No, it's not.

James Reeves (46:46)
Power BI, Tableau, whatever. So if you're quite artistically minded, do that. You could do heavily tech, which, and maybe my perspective is wrong, but it's mine, which is that Python and languages like that are tech, because you're literally coding up solutions. So my advice would be pick what you like doing.

Don't try and be across all things. That comes with time. So where I can do the sequel, can do the Power BI, can do the Dangerous Engineering, because as you mentioned, my career history, that came through time. And it also means I'm not the best at any one of those things. You can't be the best in more than one thing, really.

David Morgan-Gumm (47:28)
Mm-hmm.

James Reeves (47:33)
because there's just so much to it. And also what is the best? Who are you comparing to? I wouldn't... Even though I've just said all that in my statement, don't try and compare yourself to others. Everyone is different. Everyone is living their own version of their world. Be kind to yourself. Because you'll learn something and...

then it will be deprecated. you spent, I know one person who put a lot of effort in learning a particular tool, which got killed. And you could be really disheartened by that because it's years of practice and knowledge. But actually the learnings were applicable elsewhere.

So that would be a big piece is a lot of what you learn is not technical.

I am presented with a problem and I need to fix that problem. What you learn is that how do I fix the problem? And it's not, I'm use this tool, I'm gonna use this language, I'm gonna use this procedure. It's you learn to look at the problem and understand how much effort you should go to as well. Because you can go, well, I need to make it faster. Do you spend four days doing that?

You know, this thing runs once a year and it takes five minutes. Do you put four days work into making it run in three minutes? Probably not. cause you know, the return on investment and that is decades or centuries. so a slightly that's, that's more about looking after yourself in terms of don't, don't be disheartened and disappointed in terms of tech stack and things like that. You can, you can't go wrong with SQL.

David Morgan-Gumm (48:59)
No.

Yeah.

James Reeves (49:16)
It's such a good foundation to everything. again, it's not necessarily learning select and then group by and the order and that highly technical. lot of people new to industry I speak to, they're hyper-focused on how do I learn the language? How do I understand how to do this and how to do that? And I think the order of execution in SQL.

David Morgan-Gumm (49:20)
Agreed.

James Reeves (49:45)
It's nice to know, but is it that important? What's more important is do you enjoy it? If you don't enjoy it, try and find something you do. There's a phrase I heard years ago. can't remember. I'd love to attribute it to the right person. But I don't know who it was. And I think it's a famous one. I could probably Google it. Is if you enjoy the job you're doing, you'll never work a day in your life.

And I genuinely believe that. Yeah. Well, on a Sunday, I'm gonna call with you about SQL because I enjoy it. I find it fun. I genuinely find my day-to-day job fun. Although I said to my wife earlier about this conversation, what would you recommend if you could give someone the recommendation?

David Morgan-Gumm (50:12)
It's a motto to live by.

James Reeves (50:32)
It was that, was find something you enjoy doing. But I can't remember what the other piece was. That's annoying. There's an ADHD thing for you. can't remember what it was. It was in my brain, it was next to come and now it's I can't remember what it was. Yeah, it was essentially do what you enjoy. Oh, that was it, sorry.

David Morgan-Gumm (50:45)
I get that all the time. Just go blank.

James Reeves (50:54)
It was that for the first few years, because you look at someone like me who's been doing it for a while and you say you need to enjoy your career and cool I need to do that, but I'm decades into my career. The first five, maybe ten years, you've got to do the work that you're told to do. And you might not enjoy that. So taking advice from someone, you have to enjoy the job you do.

Yeah, but take that in context of where you are in your career. I mean, you might not have the freedom to pick and choose your projects. You will do eventually, but at first you might not. But you still know whether, I wake up in the morning and dread the fact that I'm going to be opening management studio or opening Power BI or opening whatever tool. If you're dreading it the tool, maybe try a different.

David Morgan-Gumm (51:22)
Mm-hmm.

James Reeves (51:45)
Try to pick a different direction and don't try and keep up with all the things. Just fart in like that. You just trash yourself and burn out and try to be the best of everything. You don't need to be best of everything. You need to be happy and

David Morgan-Gumm (51:51)
No yeah, there is far, far too much.

James Reeves (52:03)
effective and that's why I think the phrase effective is best. It's not the best at something, you're not the worst at something, you're not the most productive, you're effective. That's all you can ask of anyone, be effective.

David Morgan-Gumm (52:17)
Yeah, yeah, that's a very good point and a very good answer. I know that you're quite passionate about growing others and making sure that people have...

James Reeves (52:20)
Thank you.

David Morgan-Gumm (52:31)
the right, I suppose, foundational knowledge and awareness of various different things to make the most out of their job. Do you have any tips for any kind of senior people that can help them better support junior people in the team, especially within this age of major change in technology and fast implementations?

James Reeves (52:51)
Yeah.

Yeah, I would say tell them the why. If you need someone to do something, tell them why. What's the reason they're doing it? People work for a reason. Go and build this. Okay, why am doing that? I'll ask personally. If I am doing something that I don't know why I'm doing it, kind of will, well, why am I doing it? It's pointless. Sharing.

David Morgan-Gumm (53:01)
Yeah.

Hmm.

James Reeves (53:21)
the yes this feels painful or yes we did this already or if you you as senior are thinking those things hopefully you can share those thoughts of yes I know we did this already but we need to redo it because share that because otherwise the junior is just out to go in I've done this five times they don't know why they're doing it again or changing I think that helps share share the purpose

David Morgan-Gumm (53:35)
Mm-hmm.

James Reeves (53:48)
I think.

David Morgan-Gumm (53:49)
why and the purpose

is a really important part, bit of understanding in tech. There's so many ways in which you can implement something or do something, find a solution to something that you have to understand the whys to be able to come up with the best solution. And that is a hard skill to gain when you're a junior, And I remember back to when I

was straight out of university. think I was quite naive with a lot of that, you know, thought I understood the why, didn't understand the why. And, you know, you end up being a lot more helpful, being a lot more happy with the results that you provide in the end, if you understand why you're doing what you're doing and what problem potentially you solve.

James Reeves (54:38)
Let's have a purpose, let people have a purpose. We're human beings, we need a purpose.

David Morgan-Gumm (54:39)
with the work that you do. Yeah. Yeah.

Yeah, agreed.

David Morgan-Gumm (54:47)
James, this has been an absolutely fantastic conversation. I've really enjoyed it. Everything we've been through, especially the chats on all of the new diversity and equality bits is something that I myself need to learn a lot more about. So thank you so much for taking the time to share your, know, your expertise on everything from data platforms to building more human centric teams. It's been incredibly insightful.

James Reeves (55:10)
thank you thank you for inviting me on i think the podcasts are ⁓ are great so thank you having me on

David Morgan-Gumm (55:15)
Great. Thank you. Amazing.

And a huge thank you to everyone listening and watching. We truly appreciate you spending your time with us. Again, if you enjoyed this episode, please do us a huge favor and hit that like button, subscribe on YouTube, or your favorite podcast app. on loads, goes out to all of them. And share it with a colleague who might find it useful.

For more content and to check out our other episodes, head over to sqlsquared.co.uk and all the links from James and the topics we discussed will be in the show notes. Also, please remember if you have any questions to ask that you want answered on the podcast, and there might be some topics that come up in future where they fit in perfectly, please email them to mailbag at sqlsquared.co.uk. That's all from us this episode. Keep learning, stay curious, and never stop innovating. We'll see you in the next one.

Cheers, James.

James Reeves (56:07)
See you later, cheers mate, bye.