Chrissy Galloway of ENMAX and Jules Mayberry of Waste Connections went in on vibe coding, winning over payroll, and the horror of zero comments.
Emma Coppin
Content Lead, Multimedia Storytelling
mini传媒
Vibe coding. Job-shadowing truck drivers. 4,000 accidental emails. Chrissy Galloway, team lead for data and reporting optimization at ENMAX, and Jules Mayberry, mini传媒 Extend developer at Waste Connections, discuss the day-to-day reality of app development. They talk about the rush of building software that field crews actually want to use, the necessity of job-shadowing your users, and why human oversight is still required to keep AI in check.
Jules Mayberry, mini传媒 Developer, Waste Connections
Jules is a self-confessed sci-fi nerd with a passion for the technology side of HR. She is dedicated to simplifying and optimizing others' work experiences with a developer mindset for creative solutions.
She prides herself on being a compassionate leader, a fast learner, and a reliable team member who cares a lot about the success and well-being of both organizations and their people.
Chrissy Galloway, Team Lead, Data and Reporting Optimization, ENMAX
Chrissy has spent her career managing data in a variety of industries including nuclear power and oil. She currently leads Data and Reporting Optimization at ENMAX. As an avid puzzle lover, she is committed to bringing calm to data chaos and thrives in solving complex problems for her organization and team.
Chrissy Galloway: Welcome... hold on. (laughs) Okay, this is serious. Um, is this where I’m supposed to do the whole script part that I studied and then forgot? (laughs) I went from like the chaos of managing a whole bunch of requests to starting to code. You open the preview screen and you’re like, "Yes! It worked."
Jules Mayberry: Oh my god, it is a rush. We have to have a deep understanding. We have to own our solution, what it can do, what it can’t do. You can’t leave the puzzle unfinished, you can’t just leave it on the kitchen table. You want that satisfying pop into place.
Chrissy Galloway: The last piece in. Done.
Jules Mayberry: Exactly. Yeah. This is Jules Mayberry from Waste Connections. If you want to talk about building data pipelines that turn chaos into great solutions, Jules is your person. What you really need to know about Jules is that she is a former member of the Coast Guard, and now she is doing what exactly, Jules? Can you explain to us a little bit about your business?
Jules Mayberry: Uh, yeah. So, at Waste Connections, I am the Extend Developer. So, I'm on the HRIS team, and I do the app development for our mini传媒 tenant.
Chrissy Galloway: And can you tell us a bit more about what type of people Waste Connections employs?
Jules Mayberry: The best. Truck drivers, the real blue-collar men and women out there managing the landfills. And this is Chrissy Galloway from Enmax. She does data and reporting optimization. She has also worked in nuclear power and oil. Um, she also loves a good puzzle. So, she's the person you want to go to when you want to untangle some data chaos. Chrissy, tell me a little bit more about the business that you work for.
Chrissy Galloway: Uh, thanks, Jules. Uh, so I work for Enmax. We are a utility up in Calgary, Alberta, Canada. Um, so we do everything from power generation to power distribution right to your houses.
Jules Mayberry: So tell me about your team size, what kind of people you hire.
Chrissy Galloway: So Enmax has about 1,500 employees. Uh, we've got the office workers, uh, that are... so my team is finance. Um, we've got our HR, and we've got project managers running our construction projects. We've got the field crews, which are actually building new infrastructure, and then we have teams as well operating our power plants.
Jules Mayberry: So, I haven't built an app for like our actual field workers, you know, the guys in the trucks at 4:00 AM. Um, but I do imagine that if I ever have to build an app for them that they would use, it would have to be mobile-friendly because, like I said, like they're already in there clocking in and clocking out. They're on the go. They can't really be on their phone while theyre driving. Um, but then my second part is anyone that I'm building an app for, I want to get that buy-in early. Like, I don't want to just assume and make all the decisions cuz we don't have like this whole team to kind of funnel it out. So, I have to have buy-in really early so that I'm not building something and spending all of this time on something that is not going to work for them.
Chrissy Galloway: That makes sense. So, almost like have a representation of the users of the app present and collect that feedback.
Jules Mayberry: Exactly. Yeah, I call it job shadowing. Like, whoever is asking me or bringing a use case to me, okay, cool, I'm going to sit with you for weeks and just kind of figure out what this process looks like for you. And then I'll go figure out what we can do to fix it.
Chrissy Galloway: And you've done that with groups already?
Jules Mayberry: Yes. Yeah, so I got really close with our payroll team. Shout out to them, I love those ladies.
Chrissy Galloway: Nice.
Jules Mayberry: Um, and we became besties through building an app together that helps them. So, it's pretty cool like stepping into this role as a developer has actually made me more social and confident, which is pretty interesting to watch happen.
Chrissy Galloway: That is great. Can you tell me what the app does?
Jules Mayberry: It processes like bulk bonuses, compensation changes. Um, and I think the coolest thing about that app is not only does it like automate that process and takes hours off of like payroll's Monday morning, but it has like all of the guardrails and the validations and the custom error handling that you can do in Extend so that when the request goes to approvals and gets to payroll, they can trust it and just hit submit. Whereas they were taking in all of these spreadsheets and building an EIB and having to go back to the managers and saying, "This is the wrong employee ID number. Did you mean 3,000 or 300?" or whatever. Um, so it was really cool to build something that was more accurate and efficient when it comes to pay.
Chrissy Galloway: Yeah, I love the error handling part of it.
Jules Mayberry: It’s my favorite.
Chrissy Galloway: That's smart. And now that you've done that, do you feel like they are coming up with more ideas and they're asking you for more help developing?
Jules Mayberry: That's like the biggest thing. Once we got that first app out, and it was a really big app, and once they were able to see like what you can do with this product, now it's like, "Oh, we have all these other ideas now." Um, so that's why I was able to really change my role to developer, which was so cool and exciting for me. It's like, "Okay, yay, like now we've figured out what this can do and now we get to do a lot more."
Chrissy Galloway: It's really hard to envision what it can do. Cuz I've had that similar experience where it's like, "Hey, give us use cases," and you get back crickets. Like, nobody knows what they're asking for because they don't know what it can do. And yeah, so we had to put that first app out there and like give them that inspiration and then all of a sudden it was just like boom, "Can you do this? Can you do this?" and yeah.
Jules Mayberry: Yeah, when I like first built out this payroll app, I just did it with all, "What would Jules want when Jules was in payroll?" Um, and so I just kind of built out the shell and then I gave them a demo, and they were like, "Okay, cool." And then they started seeing things in their head, and they started like wanting to really be involved, and then they were making decisions, and I was like, "This is so fun, like you guys are giving me things that you want, like things that I can work through." Um, and at the end, it ended up being a giant app that everybody loves, and it works for them.
Chrissy Galloway: That's awesome. So tell me, Jules, how do you prioritize all of these demands that come in from the business? So you've you've given them an idea or maybe you've built an app for them and all of a sudden like those floodgates have opened and the ideas are just coming at you. Um, how do you manage all of those requests, and what do you decide to work on given that you've only got so many hours in a day?
Jules Mayberry: Yeah, so I'm still currently accepting any and all advice on how to organize (laughs) this. But right now, my process is I'll set up a bi-weekly recurring meeting with whoever asked for this use case, and I will dedicate that week on that app and working with them, asking them questions, and then at that Thursday meeting—we call it Thursday demos—um, I'll demo where I'm at, ask them my questions, and then by the next demo meeting two weeks later, I'll have everything they asked for and more. And we just kind of work through it like that. How about you? How do you do it?
Chrissy Galloway: Um... It's... I want to say the same as you. Like, it's overwhelming at times. Uh, we can go through cycles of tons of requests and like unable to fulfill them all to like nothing, and then we get to play catch up on our work. Um, so my team is actually uh three people. I have an adaptive planning developer, I have got my mini传媒 Extend developer, uh and then myself as the team lead uh is kind to help balance their workload.
Um, so we actually just had... we just had someone come back, an app that we had a contractor build for us last year, and they were like, "This is great, it's working well. We want to expand it now for this next use case." Um, and so my main mini传媒 Extend developer is working on another enhancement. So, last week I was like, "Well, we're a little behind schedule, so I'll jump into this one and I'll start developing through and kind of getting it to a spot where I can hand it back off to her when she's got some time." Um...
Jules Mayberry: I love that you're the kind of leader who does that.
Chrissy Galloway: I love it though. It was... it was... I went from like the chaos of like managing a whole bunch of requests to opening up that app builder and starting to code. And I don't know if you get that same rush where you open the preview screen and you're like, "Yes! It worked."
Jules Mayberry: It... oh my god, it is a rush. That's like the perfect word for it. Like, you just feel so cool. Like, "I'm doing something."
Chrissy Galloway: It does feel cool, doesn't it? Well, and then I don't want to put my laptop away and I'm like, "No, I'm just going to take it home and I'm just going to work extra tonight." Like...
Jules Mayberry: I never even thought about the enhancement piece because I'm by myself. So, it's just me kind of doing all of it, and so when they did ask for extensions to that... at the payroll app that I was telling you about, and at the same time, I'm having to research and look into building two new apps. And so it's like, it would be really nice to have a team doing...
Chrissy Galloway: It would, yeah. Yeah. But on the flip side, sometimes it's easy to manage your task list in your head. The game really changes when you have other people and other task lists to work with, because then you can't just rely on your memory or um like I'm sure you've done the same thing where you've developed something and you're like, "I'm just going to leave this piece as something I'm going to come back to later." But now you have to communicate that through to your team members that you're working with, so it does add a layer of complexity to it.
Jules Mayberry: It kind of forces you to want to like throw in the comments and make it friendly from the jump because yeah, if I were to pass this app off to another developer right now, I... I need a couple hours to throw some comments in there to explain it, for sure. How well do you comment your code?
Jules Mayberry: Uh... (laughs) I'm getting... can I ask that? (laughs) I'm getting better. Yeah.
Chrissy Galloway: It's tough. Um, I think we all do it. As in, we're all terrible at it. (laughs)
Jules Mayberry: You hear about it in all of these like um like sessions, like the demos that they do or like the program update calls and whatnot um where they talk about like, "Be nice to your future self, comment your code." And I'm like, "I will, I'll get to that."
Chrissy Galloway: Yeah, later. It's not a priority, right?
Jules Mayberry: I get it, I'm still here.
Chrissy Galloway: Until it is, until you open it up and you're like, "What was I thinking?"
Jules Mayberry: And that happens sometimes, too. Sometimes it'll be a change I made a week ago and I'm like, "Why did I do that?"
Chrissy Galloway: Yeah, yeah. So, I guess you are a real developer.
Jules Mayberry: Say it again. (laughs) So, I definitely look forward to making it easier to develop things or um making the code and the functions more easily accessible. So, we've already seen it with uh the code mode versus visual mode where uh the visual mode is more of the drag-and-drop pieces and it just builds the code for you and kind of takes that um hard to start development piece kind of out of the way. Um, sometimes I've stared blankly at the screen and I'm like, "I don't even know how to start coding." But then once I get some of those building blocks in, it's really easy to start adding the code in between. Um, but where I find this interesting is that I still don't think you're going to be able to do it without a human. Like, I think that humans are still going to need to be there to connect all the dots, to kind of be that overarching brain center. Um, I think that we still need to be able to almost structure the data in the backend. So, when you're working off of mini传媒 data, the benefit is all of your structures are the same, and so maybe an AI bot or agent will know how to communicate with those. But what if you've set your structures up differently at one company versus another? Somebody still has to translate those requirements over, and just because Waste Management does things one particular way doesn't mean those processes exist, I guess, in the same way somewhere else. So, I still think somebody's going to have to be there to control the bot, I guess.
Jules Mayberry: Yeah, like QA it.
Chrissy Galloway: Yeah, exactly. Exactly. And tell it what you want out of it.
Jules Mayberry: I have like kind of the same perspective. Like, I love the idea of using AI to make things faster, um but the fundamentals need to be there. Like, right now, when we take in that problem or that process that they want to automate or do something with, we have to have like a deep understanding of what it currently looks like so that we know how to properly solve it. We have to be aware of like the limitations, we have to own our solution—what it can do, what it can't do. Um, so I don't want to be the one who's over here just like vibe coding, using AI to build a whole app, and then when it accidentally sends 4,000 emails overnight, I can't explain where that came from or like how to fix it. So, you have to... you have to be able to take that output and review it and test it with real use cases and real problems so that you're not putting out something with lower quality. You might be able to make more apps faster, um but quality is not where we should compromise.
Chrissy Galloway: Yeah, I agree. I think that it will make our development faster for sure. I think that being able to repeat certain components over and over again, it'll do a much better job at, or like maybe it gets to the point where um you write your code or you've kind of structured your code and it just has a better way of like compartmentalizing it into um like, I don't know if you use pods or anything like that like within the code...
Jules Mayberry: Love pods. Do you? Love pods.
Chrissy Galloway: So, I don't use... I don't use them as well, but we've um my team definitely looks at how other coders do their coding and and where they're putting their scripts and stuff like that, like whether it's in the library or the um like the WQL queries are somewhere else.
Jules Mayberry: The scripts, yeah. Exactly.
Chrissy Galloway: I feel like maybe that's where the biggest benefit is going to come from is like maybe standardizing how that's all done. But kind of like what you said, you still have to peer review it. Someone still has to...
Jules Mayberry: Exactly.
Chrissy Galloway: ...put their eyes on it and make sure that it is working the way they want it to.
Jules Mayberry: Yeah. This has really been a great conversation, Jules. I'm really glad that you were able to come out and have this conversation with me.
Jules Mayberry: Yeah, absolutely! It's always a great time talking to you, Chrissy. I can't wait to bring up this conversation again after.
Report
A gap between effort and impact is the real ROI problem. Discover the solution.