Transcript
Hello.
Hello. Hello. Hello. Oh, we've got music. Woo. Well, this is exciting, isn't it? Yeah. Oh, yes.
I love it. I love that.
Love the enthusiasm. Thank you, Ellie. Welcome everybody to DevOps Dream in London twenty twenty four. Can we get a cheer? Woo.
Yes. Awesome. It's an absolute pleasure to welcome all of you here today.
It really means a lot to us as organizers of the conference to have such a great turnout and have everybody here to hopefully learn a little bit, network a little bit, and connect with the Salesforce community that we know and love. So it's a a real, real pleasure.
Of course, this event can't go on without, all the sponsors, that we have at the event. So a huge thank you to all of the sponsors and a huge thank you to all of the amazing speakers in the room as well. Thank you so much to all of you.
First a little bit of housekeeping.
We have no fire alarm expected today, so if you hear one, run, I guess.
There's a river there, if you need to cool off.
On your seats, you should have two bits of paper. One of those is gonna be the agenda for the day. All of the sessions and where you need to be for each one of those is in the blue one. And then on the I think I think it's black. I don't know if I'm color blind. The black one has the state of Salesforce DevOps, survey on there. So every year, Gearset run a report, and a survey to find out some insights from you, the Salesforce community, about how teams are approaching DevOps and their software delivery.
If you have time, please fill that out at some point over the next few weeks, take it home with you, consider some of those answers, and send them back, and I believe there's a little little prize and a little something actually in it for you if you do that. So, yeah, like I say, we can't do this event without the sponsors, so please go and visit our booths. They are all in the registration area that which you will have been to to get your badge. I see everybody's got one, so hopefully you know where that is.
And the last little bit of housekeeping is have some fun. So these conferences I've been attending for two two and a bit years now, in my current role, and there really is so much value to be had and enjoyment to be had out of it.
And the other little piece of advice, if you're familiar with Ted Lasso, is to be curious. So everybody in this room has a story and it's a unique story. So be curious, ask people about their stories, ask people about what brought them here, what their team looks like, and what they're looking to achieve over the next year, two years, three years, five years, whatever it may be, and you might learn something, that might surprise you. So be curious and have some fun.
But first, to kick things off, Salesforce DevOps, Past, Present and Future, the opening keynote.
For those of you that haven't met me before, my name is Jack McCurdy. I am a DevOps advocate at Gearset. So my journey in Salesforce ecosystem, is probably a little bit different to a lot of the people in this room. Everybody has a unique way of getting here and where they're to, but I joined Gearset coming up to six years ago now. So, March March twenty twenty nineteen, I joined Gear Set and that was my first entry into the Salesforce ecosystem as well. So I have done nothing nothing other in this space other than learn about DevOps, talk to teams about their strategies, their processes, and seen quite a shift over the last six years or so, in the DevOps space. And that's what I was gonna talk to you a little bit, about today.
The last time I did this talk, it was the middle of summer in Chicago, so I don't know if this slide really landed at all, but if you're not familiar, this is the Muppets Christmas Carol.
My boss is down the corner here and she's looking at me funny like, Jack, what are you doing? You are the muppet up on stage if you're bringing this up. Right?
But what I think is interesting so past, present, and future, my mind instantly went went to this story, and the reason for that is if you take a look at the story, we have a journey. So Ebenezer Scrooge meets the ghost of his Christmas past, his Christmas present, and the Christmas future.
And across the story, what we find at the end, Ebenezer Scrooge had learned some hard lessons and some hard truths about himself on reflection from these ghosts of Christmas past. And a lot of the conversations that I've had in the ecosystem, many people think that their future outcome is predicated by what has happened before. You know, we've always done things this way, this is kind of the path we're on, and this is where we're gonna end up.
And we've come to learn that that actually isn't the case. Much like Ebenezer Scrooge, he changes his behaviors, and he's able to alter his future. And that's what we can do today by taking in all of the lessons, by connecting with this community, and learning a little bit. So, yes, well, initially, you saw the slide, Jack, you are a muppet. That is why I think this story really resonates with where we can go and where we have the potential to go.
Also, it's nearly Christmas rather than the middle of summer, so, I don't look quite as insane. Who here has been personally victimised by a change set?
Yeah. Yeah. Nearly nearly nearly nearly everybody. So when I first came into the ecosystem, change sets were kind of the default, the de facto, especially amongst Salesforce admins used by developers when they needed to as well.
Change sets, as we've all seen, have the opportunity to really frustrate us, to really create challenges across the delivery life cycle, especially when you spend so much time, so much effort, so much hard work building great features, you get to the change set stage and you feel like you're being slapped in the face like these unfortunate admins, developers, and release managers here. But that was very much one of the two camps that we found in the past, these people that use change sets.
And then we had the development side of the table, the people that were more familiar with using tools like Versus Code, the Salesforce CLI, and GitHub. So traditional development tools that many software engineers that might have transferred into Salesforce development were familiar familiar with and comfortable using.
So we have kind of the camps, the declarative declarative admins, we had the developers, and these siloed teams and these siloed people within these individual teams, there's been friction in the past. You know, five years ago, it was always our admins have deployed one way, our developers have deployed another. We don't really have any visibility across those changes and who's made what and those processes are not unified.
The good thing is is that that's not really the case anymore. So that State of Salesforce DevOps survey that I just told you about, the results from twenty twenty four showed that most teams now are deploying in a unified fashion and some of the reason for that is is that there has been a lot of evolution across the ecosystem.
There have been vendors like Gearset, Capado, Autorabbit, you name it.
These tools have evolved born out of the need to put better tooling into the hands of our admins and also our developers. I'm not saying that, the developers, absolutely nailed it, but their processes are rooted in software delivery best practice, which is, of course, been established for a long, long time.
So this evolution is really important about how we are kind of at the stage where we are now. We have a lot of powerful tools at our disposal and we have a lot of concepts that are now more familiar with us, you know, this is DevOps streaming, you know, five years ago, did we expect to be in a conference room talking specifically about DevOps for a full day? Probably not, but that's where we are now.
And that is to say that in the past, we've been mainly focused on the actual deployments. The deployments is what has given us the biggest pain and the biggest challenge across the ecosystem, but that's not necessarily the case anymore. Focusing on deployments is fine. It's a challenge. That's why a lot of the DevOps tooling that has come out, come out over the last few years has existed is to solve this specific challenge.
However, where we are now, we're now concerned with more of this loop, more of the facets of DevOps that are incorporated, and you'll see that across a lot of the talks in that little blue booklet that you have today. Many of the talks encompass much of this life cycle.
Things like observability are more important, testing is more important, backup is more important, and we have tooling now coming into the fray to support some of these things.
But it's been a really interesting journey for me personally to kinda see all of these tools come to light and see just how mature the ecosystem has become over the last five years because I really think that it is testament to everybody in this room and the wider ecosystem generally that we are looking to do the right things. We're looking to not only build the right things, in the first instance, but we're looking to manage that process holistically. So that's where we've kind of come to now. We're more aligned to this DevOps infinity loop, if you like, than we have been previously.
You're probably wondering what's the what's the reason for this? So if you cast your mind back, I think, twenty twenty or so, Permageddon. Does anybody remember that fateful event? We got we got a hands up. Yeah. We got we got a couple of couple of people in the room.
So an event like Permageddon. So Permageddon was an event in twenty twenty, early twenty twenty, or early twenty mid twenty nineteen, around around that kind of time.
Salesforce pushed an update, which gave admin access to all of their users on specific instances of orgs that have Pardot.
So people were logging in one day and they realized that people could edit anything.
At this time, very few Salesforce customers had robust solutions in place and were running around with their hair on fire with all of these permissions and had to lock everybody out. So Salesforce put in a fix where they locked everybody out and nobody could do anything for days.
The lack of robust backup solutions, observability solutions to identify these kind of problems led to users logging in and having access to all kinds of data that they shouldn't have done, and a real head scratcher for a lot of folks.
And this isn't too this this event in in of itself is a bit of a one off, but it's not too uncommon across most of your teams. You deploy, permission that isn't isn't quite right, it ruins access, or you try and deploy a profile, try being the operative word because they're a nightmare, but you deploy a profile and things can go wrong pretty quickly. So that's why we've seen the rise of these backup solutions, these monitoring solutions, these observability solutions, and why that we're more concerned with more of that loop now because we've seen just how wrong things can go.
Well, if we take a look wider, just beyond the scope of our deployment needs and our DevOps needs and those practices, the ask of us as technologists now is so so vast. Five and a half years ago, I think I spoke to people that believed they did know most things about the Salesforce platform.
I won't do a show of hands because none none will go up, but I can't don't think anybody in this room or any room in a Salesforce conference can confidently hold their hands up these days and we so we know every single part of the platform.
Might have been the case, a few years ago, but there is so much more coming into it now.
Salesforce platform, data clouds, all of the customer three sixty work, Agent Force, and all of the different clouds have their own nuances and have their own specialisms.
And our job as technologists is to specialize now in these fields and in these specific parts of the platform and deliver value to our business.
And how do we deliver value to our business? We go back to our change sets and we're trying to deploy these great new flows or great new I don't know if it was process builders back in the day.
Great new customizations and deploy customizations to different clouds and that was a challenge.
So DevOps is rising because of the needs of the platform are greater, because your job is stressful enough as it is without worrying about the deployment stage.
So we're really seeing this seismic shift in the way that we go about these, the need and desire for teams to have these robust processes are in place, and it's really it's really exciting. Don't get don't get me wrong. I am excited to see how people use Agent Force and how the different clouds develop and really see what we can do. But to support all of that, we need the robust practices in place to be able to do that successfully, successfully, to deliver that value.
And also not not not to stress yourselves out. Right? You wanna focus on building these things. You wanna get your hands dirty. You wanna get into the weeds, and if we have systems that support us, then we're gonna be happier and more confident in doing that when it comes to the deployment stage so that we can focus on delivering that value and learning new things.
Wouldn't be a keynote without talking about AI, would it?
AI is is here to stay, you know, that is free.
My views do not necessarily represent that of my employer.
AI is here to stay, I I believe.
Gearset recently sponsored this DORA report, twenty twenty four DORA report. So, the DORA report is a paper that is released every year, by Google Cloud's DevOps research and assessment team. A lot of that report this year is focused on AI and the impact of AI. And seventy percent of the respondents to that survey said that they rely on AI for at least one part of their job. Who uses AI day to day in some way shape or form? Yeah.
So this thing is not going away. You know, the actual definition of AI is a computer's ability to execute tasks that might otherwise have been performed by a human. And if you take it by that definition, AI has actually existed for quite a while. But we need to be really specific and really conscious about what we mean when we're talking about AI solutions because if I say AI in this room and most other rooms, we're thinking generative AI and large language models.
I'm not gonna go dive too much into it. Peter Chitam is actually doing a talk on large language models later, which he did another conference I went to, which is brilliant. So, to learn more about the actual implications of this, definitely go and see Peter's talk.
But this is this is here to stay, and we have some really key considerations about AI and how we are gonna use it moving forward.
And some of those main considerations are actually where do you employ it into this process, where do you see value in the process of adopting AI based solutions, And what is actually the definition of AI for for you and your team, and indeed for potentially any any vendors out there that might be using AI capabilities? What does it mean and where does it add value? Those are some of the biggest considerations I see. And where should you actually be using it in your process? Should you be using it to help plan your releases? Should you be helping it to generate code for you?
Should it be in any of those kind of observabilities? What kind of level of insight is AI gonna give you? And that's what I would encourage people to look at is if we are using something that is AI based, what actually about where where is it valuable, and how are we actually applying it? Why are we using AI just because it's AI and it's the fancy thing? Always focus on that value.
I would a word of caution. So that DORA report that I just mentioned from this year, though respondents said that they use AI for at least one part of their daily job, one of the main things that we're concerned about DevOps is our release velocity and our release stability. So there are four metrics in DORA that cover those two, the split across those two things, and implementing AI solutions is actually driving down the velocity and the stability of those teams that are choosing to implement AI solutions. So those are some really considerations that you really need to take on board if you're gonna do this successfully and if you are gonna travel down that path.
So cautionary tale and more on the Dora report. If you want to read that, I highly recommend that you do dora dot dev and the report will be there to download some really interesting insights.
But something else that the Dora report also said was that the developer experience is the key to most team success and happiness.
Talked a lot about tools, technology, the things that help us get to where we are, But to me, the future of Salesforce DevOps is in the people.
The future of Salesforce DevOps is human centricity and how we bring our people together to deliver the best solutions possible.
That developer experience and that human experience that we all feel is gonna be really important as we continue to consider our processes. How do processes changes affect our people? Can our people deliver those process changes? Can our people deliver the technology and the value that we're going to? And you'll see a lot of that, I think, today across some of the sessions, some of the connection that we have with our teams. If you look at the job market these days, how many jobs who's in a remote job?
Remote? Yeah. Most hands. That's why sessions like Andrew, Andrew Cook at the back will be doing a session later on digital body language and how we communicate with people across these new platforms and these new way of workings.
Who likes working remote? Yeah. Right? Me too.
So these considerations and this blended world that we're now living in, that human focus on our teams, how do we make sure that a team has everything that we need? Do we go down a platform engineering style route and give them a certain set of tools that they can self serve and use to get the job done?
Another big topic of the Dora report, but I'm just giving it all away. You won't need to read it after this talk.
Really important considerations is how we interact with people and how we take our DevOps to the next level is within those connections.
So that's what I mean by stay curious and that's what I encourage you encourage you to do at the start of this session is stay curious. How are people working together? How are they bridging that gap? How are they feeling about their processes, about the technology that we're using? And I think the more that we look at the people side of the process, the more that we're gonna uncover not only about ourselves, but where we can take the technology to get us there to the right place.
I don't have too much more to say, so that is that is what I think and I think again, it is testament to all of you being here just how big this topic has become, how important this topic has become, and how has become, and how much value can be gained if we dig a little bit deeper, if we dive into some of those sessions that we have today, whether it be better data management, whether it be how to implement a DevOps process for large teams, all these micro considerations will have a huge huge impact on not only your success, but your happiness and where we go forward.
So with that, thank you so very much and if you have any questions I have, I have five minutes, so thank you.
Stand silent. So what happens what happens now is we have a short break, until somebody with an agenda tell me quickly, till ten thirty. So go and grab coffee, start some of those conversations that I mentioned, and really plan out what you're gonna get out of the rest of the day, those ones that are gonna give value, and have a great experience. I'll be around all day. Feel free to pull me aside and have a chat, and have a great day everybody. Thank you so much.
