← Scrum Master Toolbox Podcast: Agile storytelling from the trenches
Scrum Master Toolbox Podcast: Agile storytelling from the trenches
BONUS Building Reliable Software with Unreliable AI Tools With Lada Kesseler
00:39:08
Loading summary
Transcript59 lines
- [00:04]
A
Hey there, agile adventurer, just a quick question. What if for the price of a fancy coffee or half a pizza, you could unlock over 700 hours of the best agile content on the planet? That's audio, video, E courses, books, presentations, all that you can think of. But you can also join live calls with world class practitioners and hang out in a flame war free and AI slop clean slack with the sharpest minds in the game. Oh, and yes, you get direct access to me, Vasko, your Scrum Master Toolbox podcast. No, this is not a drill. It's this Scrum Master Toolbox membership. And it's your unfair advantage in the agile world. So if you want to know more, go check out scrummastertoolbox.org membership. That's scrummastertoolbox.org Membership. And check out all the goodies we have for you. Do it now. But if you're not doing it now, let's listen to the podcast. Hello everybody, welcome to this last of the week episode on AI assisted coding. And to join us for this episode we have Lada Kessler. Hey Lada, welcome to the show.
- [01:24]
B
Hello. Thanks for having me here.
- [01:26]
A
Absolutely. So Lada is a passionate software developer specializing in the design of scalable robust software systems, which I'm sure is going to be a topic today with a focus on best development practices. She builds applications that are easy to maintain, adapt and support. Lada combines technical expertise with a keen eye for clean architecture, sustainable code driving innovation in modern software engineer. And as she puts it herself, she's currently exploring how these values translate to AI assisted development and figuring out what it takes to build reliable software with, and I love that part, unreliable tools, because that's exactly what they are since they are stochastic parrots as the name goes. So to get us started, Lada, tell us, what do you define as vibe coding and how do you compare that to what you're exploring now? Other approaches of coding with AI.
- [02:26]
B
Right. I don't love the word wipe coding. It kind of became a little bit of a dirty word. Like the real developers don't want to be caught. Wipe coding, really. But actually I found that there's time, place, like it requires judgment and this, like it's never been as requiring judgment as at these times. So I think about this as one extreme on a new dimension. We used to live in a world where we could only develop software one way. And the way was you understand the problem, you go to a computer, you type some code, hopefully you do it with tests and outcomes. Reliable software, yay, for the most part. And then this year it's really, really changed. Like it feels for a long time are now different. Right. And vibe coding is basically the extreme of taking kind of that dimension to the far, far edge. Right? You just talk to a computer, you say do this, do that. You don't really care about code. Like you don't just not look, you don't care. Like maybe you don't know how. And that's fine for some systems. That's fine. Like when the problem arises is like when you put the stuff to production and you really care about your customers. Please, please don't do that. Bad idea. Like we have horror stories about people deleting databases. We have horror stories about, oh, not.
- [04:05]
A
Just any database, like the database that ran the business.
- [04:08]
B
Yeah, that's dangerous. And like you are in danger of like really losing the trust of your people of people you care for and damaging them irreparably. Like seriously, no bad idea. But also I see also the different extreme interestingly this year, so I've been holding this values that you mentioned, right. So close and dear to my heart. I really care about reliability. I really care about test driven development, evolutionary design and hisawan architecture. I do this a lot, I practice this a lot. I factor a lot. And my systems work silently. They don't make a lot of noise because they don't break. And then I see that I've been in this situation where I was like last year I was like this AI thing doesn't seem very helpful. I mean, yeah, it's in the browser, it spits out a code, it doesn't even compile. Oftentimes it's nonsense. Why would I use this? This is stupid. And then something really changed to me in November.
- [05:13]
A
What was that like? What really changed for you in November? What was going on? And then what happened?
- [05:19]
B
Right. So I was in a situation where I just developed a backend. So I'm backend developer for the most part and I did something new that was like this agentic system with rag and stuff and I've never done it before. I was so excited, I wanted to show it off. Right. And it's an internal project so it kind of naturally gets less attention. And I was trying to get a front end developer help me. So they already had started to work before me. They didn't have the backend. I did the backend but now there's this front end in angular and I don't know, angular. Like I'm. I'm kind of getting full stack but not in Angular, I can do react, but like, you know, new stuff. So I'm like, oh man, I'm stuck, right? I'm blocked. What do I do? Everybody's busy. And I remember those. Interestingly, it was pre authentic, right? So for everybody else, I think it changed at the moment. They were introduced to agentic coding. I saw it on projects. So Anthropic introduced projects somewhere in the summer, I think.
- [06:24]
A
Yeah. Project is the feature of 2024 Cloud Desktop, where you have a bunch of documents and chats that are all accessible from any chat, right?
- [06:34]
B
Not just that, but also remember, it's like, sorry, imagine it's like an encapsulated context so you can tell it, I'm trying to do this project thing, this is what it does. And then you can give it context, you can give it files on the side. You can be like, here's the file. Now it has more of your context, so it can be more helpful to you. And what I discovered was I could actually. So there was a plugin I was using in IntelliJ, it's called DevOps Plugin. And it was trying to attempt to be like this chat interface next to ide. But I've never used it this way, which is hilarious. I used only one feature of it and that feature was golden for me. What I could do is I could put it click on a folder, right click and say, grab this to my context and everything underneath everything, all files get concatenated into a giant string and then there is a menu on top which files there. And then each of the files is included as a string one by one. So it's a giant string of code. And that takes me a second. And then I can give it to projects. And unlike when you try to pull a folder, click one by one, you just give it as one file, like one thing. Now I can give big chunks of my code base to AI and that changes so much that makes it stupid and unreliable to. So I tried to do this. So I gave it my existing Angular project. It wasn't very big, so it fit. And I was like, can I even try doing this? Can I even try attempts doing front end? And I did. I was like, how would I do this? I did it like I normally would. I'm like, okay, let's start with a simple thing. Let's make a page that does nothing. Can you add a page? And it did. And I kind of like the maximum the knowledge of Angular I had was like, usually the component is like four files. CSS, JavaScript, some other stuff. So I give it the thing and it produced the page and this works. Okay, well, okay, let's add a text to it. And I added the text. Let's add a field and so on. Right. So I built a whole front end port for myself this way and it worked. What was hilarious is I got compliments from my colleagues for my design skills.
- [08:56]
A
Because I also fixed all of those.
- [08:59]
B
Yeah, because I also fixed all the little nuances, like added some pretty designs and so on. Right. And. Oh, I told them, but absolutely told them that it was AI driven. But like. Yes. So I was like, oh my God, I don't need to.
- [09:14]
A
You kind of got a new skill for free. Right? It's like suddenly it was there.
- [09:19]
B
Yes. Yeah, I can use it. And because I have engineering skills, I can understand enough, I can ask questions. Right. When I. When I'm like sussed about something and I can actually produce something reliable by doing. And I was working very closely. This is pre Agentic. So it spits out a file to me and I copy it to my clipboard and I have to go and edit the file myself. And so a little bit different than Agentic. Agentic. Now, basically you open whatever you're using either the CLI tool or the IDE and it can do stuff for you. So it's more efficient. But this is the moment that changed everything for me. Not Agentic for it. Agentic is a nice sweet thing on top that makes things much better. And also the fact that it's in your computer now. Right. So previously in the browser, it's like locked in the browser. It's like a prisoner there. But now you bring it. Agentic brings it to your computer and now you can use your tools and you're so much powerful because of it.
- [10:19]
A
Yeah. And Agentic, just to clarify for those that might not be aware, is tools like cloud code or Codex or Cursor or Windsurf or whatever people might be using that actually are there. Of course they still use models in the cloud, but they are there in the computer. So they can run shell commands, they can open files, they can edit files directly and we interact with them and they actually do the editing. So it's called agent decoding because they call this CLIS Agents. I don't think that Agentic here is the right term, but let's not discuss semantics. But one cool thing about this story that you shared is that you started as a skeptic. You gave it a shot for something that you knew you wouldn't be able to do it easily and quickly on your own. And then suddenly you got a new skill, right, with the help of AI. But you got a new skill. And I think that that is also an empowering realization that for everybody out there, that if you're facing a problem, AI might actually be able to help you solve that problem with skills you don't have yet, but can get.
- [11:34]
B
Absolutely. And sometimes even it can uncover things you didn't even know existed. Because. Do you want another story? Right. So in July, no, in June 2025, I switched to cloth code from Windsurf. It kind of switched in May, but I really started doing it in June. And this is new because this is like I was before I was in Windsurf. I was so torn because I love IntelliJ. I absolutely love IntelliJ because it's the best tool on the planet for refactoring. And I do automatic refactoring. I can do crazy stuff with automatic refactoring. Like, they dropped me into 50 microservices that I didn't know. And months and a half after, I emerged from it by doing tiny refactorings. And I changed the architecture to support the feature and also improve the architecture overall. Tiny little refactorings just rely on that and my test skills and so on. So I love intellij. Nobody on Earth, if you asked me whether I would use any other idea, I'd be like, I met. And then I switched to Windsurf. And that felt like a betrayal. And also it was annoying because I still use IntelliJ for this stuff. But then there's a gentic experience in Windsurf that's better for me. I started doing Windsurf in December and it really worked for a while. It was worse, much worse than Claude code. So when I switched to Claud code, basically this was a different thing. And suddenly I. I was in cli. When you cli, the natural thing you want to start doing is you want to open several of them at the same time. Because really, what hit me when I was using windsurf and IntelliJ was is this in my way. The IDE is in my way. It's not helping me right now. I really want to just open the folder. That's what I want. I want and I like AI in my folder. I don't need. I already have it if I look at all the good, right? And so I started opening it, like several chats at the same time and so on. And when you look at them, like, there's two Clock codes, Right? Working maybe in the same folder. I kind of naturally want to make them talk because that'll be so cool. Because for one, like, it would be nice for something else to give feedback to this thing. So I was like, can I do this?
- [14:07]
A
You had two cloud code instances in the same folder and you were just about to make them work together.
- [14:14]
B
Right. I wanted to see, can I make them talk? Because that will be so cool. The way I approached this was as I went to plot code and said, hey, cloud code. Can you actually talk to another cloud code? This was interesting because it said no, And I didn't realize. Then I was like, okay, no, sure. Like, it researched the skills like the cloud code had. All right, too bad. What can I do instead? What if I like, wait, I know that I can have it launch the script and the script can start another code. Like, can I go this direction? And I started going this direction and I had it launched another code in a different instance. And there was a mistake that I make accidentally. Like those two panels on the right side, it was popping up the new one. Then I made it double. I double clicked and I had it try to create a new one when it was already open. What was interesting, it went to the prompt part of this thing. Then I suddenly see a message from here going to the prompt of the right side. Well, that's talking to each other. That's what I want. So why on earth did it not tell me that it could? And that's when I realized the thing that I kind of made into a pattern or an anti pattern. I think it's called answer injection. Basically, you're limiting yourself without knowing you're limiting yourself. Because when I was like, hey, I see that this is possible. How can I? I reformulated the problem. I once did it to them and said, I have two panels, panel on the left and panel on the right and the terminal, and I want to send on a Mac a message from here to there. How do I do that? I described a little bit more of my form of space. Then I was like, oh, yeah, sure. Well, you can use tmux. And also you can do applescript. I didn't know applescript existed. Applescript is a crazy language.
- [16:16]
A
It is totally crazy. Yeah, yeah, yeah, yeah.
- [16:18]
B
It sounds like. Like it's English.
- [16:20]
A
Like you were kind of just having fun discovering all of the possibilities there.
- [16:25]
B
Absolutely. And it introduced me to AppleScript, which I didn't know. I didn't know. Right. So this is kind of using, like, don't limit yourself by how you formulate it's so hard to detect. You see how hard it was in this situation. I didn't even realize until this moment that I was restricting myself. AI has crazy powers. And one of them is, it's a very, very well read thing. It read the whole Internet and a lot of Internet is bad, but also.
- [16:55]
A
Like some of it is good and useful.
- [16:58]
B
Yes. And it can tell you things if you're asked the right way. So learn to ask the right way. Right. This is like one of the powers this year that's been kind of super cool. Yeah.
- [17:08]
A
Learning to ask questions is incredibly powerful. Also with human beings, but of course also with AI. Right. Like just ask questions, be in that role of the. The beginner.
- [17:20]
B
Right.
- [17:21]
A
Beginner's mind. Like, I want to do this. I don't think, I don't want to think whether it's possible or not. I just want to do this. How do I do it?
- [17:30]
B
Exactly.
- [17:30]
A
So how did you solve that then? Did you kind of just use a script to kind of send messages from one process to the other? How did you solve that? Cloth coding? Talking to another cloud.
- [17:42]
B
Yeah, so I didn't restrict myself to cloths, so I went with AppleScript. I was thinking about it. I know TMUX can talk basically on any system, Windows, whatever, Linux, tmux. But that restricts you to launching this thing in tmux. Which means when I'm dealing with Windsurf, I can't talk to windsurf. So AppleScript allows me to talk to Windsurf. So I developed this little. It's a series of small tools that together they form a toolbox for me. And this allows me to have several AIs on my machine. I could have any AI with machine talk to each other and work together if I want to. And I've been playing with it a little bit. So this is, I think later it was possible.
- [18:25]
A
Give me a new example. Like how are you using multiple AIs on the same machine right now?
- [18:31]
B
So I started trying to do something that later emerged as swarm programming. And I don't use swarm programming. I tried it. It's a lot of work. And I'm sus of the people doing this the way they're doing this. I can explain why. So just to explain, basically, let me just explain what I did when I developed my toolbox. Now I have basically I can open several terminal windows next to each other and I open clock code in each of them. And what I did. One of the parts of my toolbox is I Can name a clock code instance. And what is helpful usually is given the name that kind of shows the role. So I have maybe an architect. I don't love those roles really as much, but in this thing it was helpful. Then I had a developer and I had a tester. I did two things. I basically have a command that I can give them very quickly to say, hey, we're a team of human and clock codes. We can see who's around. By doing this, you can look at your own name and the first thing it does, it kind of runs and checks. Oh, I also see those fellows over there. Cool. What do you want me to do? Human? And then another thing I do is you also put files with description of their roles and what excitations are and when to call other things and also what they need to do. And then you just relax, tell them the thing and watch. And when I looked at it, I was like, oh my God, this is a swarm.
- [20:17]
A
That's real agent encoding, by the way. Yeah, that's what real agent decoding really is. And I'm glad you shared that example because having only one CLAUDE code, although CLAUDE has already developed and released the idea of agents where you have one cloth that can then spawn different agents either in parallel or in C ones and then kind of this is one way in which they restrict the context so that agents can focus on one specific task, which is quite similar to what you just described, but inside one CLAUDE instance. But then when you have multiple instances and you called it swarm programming, so where you have multiple AI agents doing work at the same time, I think that's what, in my mind, that's what we should be calling agent decoding. Right? Like agentic in the sense of direction and agency. Right. Wanting to do something that that's what your files description files do. Right. And make it want to do something and then putting them all together. And I think that's a beautiful story. By the way, have you published that? That set tool set publicly should probably do that.
- [21:32]
B
The thing is like, when I discovered, I was like, oh my God, my mind is blown. This is crazy. What? Like nobody talked about swarm programming at the time. And I was like, I should publish it, I should make a video. But then life caught on me.
- [21:45]
A
Now you're doing this podcast and now it's out there, everybody's going to hear about it.
- [21:49]
B
Yeah, I could absolutely release that. It is fun. I don't use it this way all that much because what I found is I need something on top. I need something that guides Them better. Because what I want to do when I code with it is I need to do micro adjustments to what's happening a lot. And I tended to go back to this one instance and talk to it. And then those are a little bit like, they still help with context. So I kind of postponed this idea a little bit. But I have been using a different thing. So what I'm doing now right now is basically I often have several, several instances and I not necessarily have them talk to each other. Sometimes I do, but I kind of check themselves against each other. So somebody has a saying of this I really believe in and it's true. I think never trust the output of a single LLM because it's very interesting and subtle. So when you ask, hey, develop a feature for me and it develops a feature and it produces a thing, then you can even ask the same thing, look at the feature, look at the code, understand the feature, understand the code, find the issues with it and give me the issues and suggest improvements. It suddenly finds improvements and it's all silly code. I do this often with two clause or with codecs lately and have them give feedback to each other and self correct because it's really a bit like tiresome to like be this person that is like hey, you know, shouting at it like no, don't do that, don't change direction anyway. And I often have have like a little committer on the side. This, this job is only commit to commit and it's just like a tiny little thing that I'm absolutely adoring. I have stories about this. So the little ones, they talk to each other. I think to go to real swarm thing, I think I need to do something else on top for myself. Other people have been doing swarms and they swarm. Sorry about them. I tried cloudflow and that's a system developed by somebody and I have a litmus test. So I have this little application that I already formulated the requirements for and I often run it with different AIs to test how they do. Like when a new model comes out, I'm like, oh, go develop this thing. And I watch, right? Is it good? Is it not good? And I tried it with Quad flow, it was the worst one. And I'm like, if you can't do this silly little thing, if I have to rescue you three times over, that why do you think the big one.
- [24:39]
A
That actually raises an interesting topic? Because of course swarming is a different problem from developing code, right? Like if you were hypothetically able to keep all the context in your Mind as a person or as an AI, then it would actually be much easier to just do it one person on their own or one AI on its own. But when you put multiple people or multiple AIs working on something now, suddenly you have shared context that is constantly changing and not always being shared with each other at the right time. So that presents a completely different problem which we humans have struggled with for many years. Right. The whole idea of trunk based development created such a controversy in our software community because it's a completely different approach to developing software when there are multiple people involved. But trunk based development was about making it easier, easier to share context, making it easier to keep process going, because that's what we learned actually works. So I think that the swarm programming really has this kind of open problem space, which is how do you share the context without overloading, which creates confusion, but enough so that the agents actually help each other rather than trip over each other.
- [26:03]
B
Yeah, and I saw a lot of tripping. Some of them have like memories like in database and all of that, claiming that it's solid. I haven't seen it, but there's people who say that it's really helpful. So I, I haven't gone all the way and tried it really, really well yet too.
- [26:21]
A
But these are, I think it's absolutely right. These are lovely stories of your discoveries and your experiences with AI assisted coding based on what you've learned so far and even your own experiences. How do you see this evolving over the next couple of years?
- [26:39]
B
Oh, I have no idea. I think we haven't even touched the possibilities of what is there out there right now. Because the ease of how easy it is to just discover new things, it's just so, so easy. Like somebody said, like we're in the era of gentlemen scientists. Like, you know, newbies can make big discoveries right now because like nobody, nobody knows what it is. First of all, the creators of LLMs didn't expect this stuff because the stories in AI engineering book is like in something like 2017, they're like, yeah, we started throwing a lot of data at this thing and it's getting better. Oh my God. They didn't expect it. The creator of cloud code, that was accidental. This is my favorite tool. It was my favorite tool. Also the creator is like, I don't know what I'm doing necessarily. I'm figuring it out. This is so new and predicting or we even don't know what this is because I think many people have conceptions about what it is and they're like oh, it's not doing perfect engineering. So it's useless. No, it's not. I'm using it a lot.
- [27:55]
A
It's not doing perfect engineering yet because of course there are these frameworks coming up like vmad Spec Kit and others that are trying to create the correct box. We have an interview with David Dahl a couple of days back. If people want to go and listen, I'll put the link in the show Notes where he presents his aid or augmented AI development framework, which he says, you know, it's about applying great software engineering practices to coding with AI. Right. So it is true that coding with AI still is a little bit hit and miss, right? Because it's not reliable, but it's only going to get better from now on, I think so.
- [28:39]
B
And also there's one question that relates to your question that is in my head that I don't know the answer to. Right. I had some earlier questions like, is code quality even important right now? And my answer is actually resolute, yes, because this thing is a monkey. And if you put it in a good code base, like any developer, it's gonna replicate what it sees so it behaves much better in the better code base or refactor. The one thing I don't know is can it manage complexity? Right now the biggest problem of it replacing anyone is if I just let it do things, it'll just run itself to the wallet, creative speeds because it's really good at running. So I have to be there managing complexity for it. Right? Doing refactoring, like refactoring to hexagonal. I do a lot of. I produce, I don't quite buy code, but I'm close enough, I'm high enough, far away enough from the code. It's a dimension, so I can be closer and further so I can pay more attention and let know. And then I go and refactor when I. When I'm like, oh, we're in danger, let's do this. And I can use. I've been using all my skills as a developer to do this stuff this, every. All the skills that I have are so useful because I know how to do hexagonal. I can tell it do hexagonal and teach it and guide it through it. And I can make a file that shows how to do hexagonal and give it to it next time. Right. But the one question I have right now about this whole thing is can it manage complexity? Can we teach it to manage complexity? I don't know the answer to that.
- [30:19]
A
And you know what's interesting about that question, because it's a question we humans struggle with as well, right? Because complexity in code bases is a problem for humans also, not only for AIs, we also have a very limited context window in our brains. It's just that we have other tools that we haven't yet been able to either replicate or improve for the AI for its own development. So I think it's really important for us to ask that question in general because these tools are becoming extensions to what we do. Just like what you described now, right? Like the way you described it is that it does stuff and then you're at a certain high level, but sometimes you dive in and you do some refactoring, whether with it or despite it. You know, I'm sure that that will depend on the problem at hand.
- [31:11]
B
And that really, really depends on what I'm dealing with. Because I use my judgment as a developer, I'm very responsible. I don't put five coded stuff in production. I don't recommend anyone doing that. So, like, when I'm working on a project that doesn't require. This is where I'm like, people are missing out when they're stuck in one level of speed. Because if you're like me, if you. I was like, oh, quad quality, always. I started seeing limitation of this this summer because you missing out so much. You need to. It's like when you're driving, you kind of need to be attuned to environment. Like, you can go fast and slow and sometimes going slow is bad because if you're on the highway, you're going to get hurt. Right? Same thing. Like depending on what you're dealing with, if you're working with something that doesn't need to be reliable, if you're learning, especially go fast, don't pay attention, it's fine. You know, you have skills and you can get back yourself. Right. What we did on Hack for Good is perfect example of this, right?
- [32:13]
A
Yeah. Hack for Good is the episode with Llewellyn Falco. I'll put the link in the show notes. Go ahead.
- [32:18]
B
Right. Because we. In three days before it, we had conversations before this, we had 10 people, Woody, Zoom, a few others, and they were like, oh, we're not going to really release anything in three days. Which means, like nonprofit people. And we really want to. But like last period, it's three days, right. And then we did. And we did two things, like, two things that nobody ever even aimed for. And I think the second thing is purely Llewellyn Falco. If Llewellyn wasn't there trying to. So like he dared, first of all, and then he managed risk to a degree that was just beautiful to watch. Right. And we used AI and we used AI and we didn't pay attention and it wasn't important because at any point we could actually step back and refactor. We weren't unsafe.
- [33:04]
A
Yeah, that's a very good way to kind of wrap that story. Right. We were not unsafe. That's a beautiful way to put it. Vlada, we're almost at the end, but is there a resource, some kind of, you know, a book, a website, a blog post, a video, something that you think people could benefit from, learn from if they're interested in AI assisted coding?
- [33:28]
B
Yeah. So help me understand. Can I give you links?
- [33:30]
A
Yes.
- [33:31]
B
Okay, so I'll give you links afterwards. So there's a. I recently did a talk on this, right. I was like. I was asked to do a talk at Calgary Software Crafters about how I code with AI and it was really, really tricky to get this out of my head. But in the end, what it turned out to be is a walkthrough patterns. So I have those little individual things that I found valuable. So I have patterns. I also have anti patterns and I have obstacles. And the obstacles are limitations of AI that the whole aim of this talk is kind of give you the foundation and specific skills. And what I was told is that this talk is really good at taking somebody from really beginner level all the way to far more advanced, which is nice because that's what I was trying to convey. So I kind of distilled my experience in a year into that talk. So there's a video at Calgary also there's a website, because I kind of had my patterns in markdown as well, and I shared them. And then Yvette was like, oh, let's make it into a website. And she did. And then I added to it and it's still work in progress. But there's a website. I made my map interactive so you can click at the individual pieces. That was very fun to do. And you can look at that. So there's also a substack if you want. I have some blogging. Yeah. So that's for myself. If you want other recommendations, let me know.
- [35:00]
A
Well, we'll put the link to all of those in the show notes. Thank you for sharing all of that with us and. And the rest of the community. This. We are all discovering this together. And that's an important thing to recognize.
- [35:11]
B
Yes. And that was the topic of my whole thing. Like we're in new territory, we're discovering it together. Here's a map of my discoveries.
- [35:19]
A
Exactly. And sharing is how we do it. And that's what we love to do here on the podcast. Sharing is caring, as I often say. And how about you, Lada? Where can people go if they want to connect with you, know more about the work that you're doing?
- [35:33]
B
Yeah. So LinkedIn would probably be the best resource. So I'd reach out there. I have a little bit of like, so I have the blog and I think I'm gonna write more soon. And a little Bull sky, but also not, not very populated right now. Yeah, but I think link LinkedIn the.
- [35:51]
A
Best way and we will put the link to that also on the show notes just to make sure.
- [35:55]
B
Also, I'll also be in DEFCON in New York on November 18th. So if you're around New York or if you want to do this online, I'm going to do this whole talk, the same talk. I think it'll be a different talk by then because a month is a century. So I'm going to do this. It's a workshop. It's a two hour workshop and I kind of walk people through the patterns and I think I'm going to add a few more and tweak it a bit as well.
- [36:23]
A
Very good. And we'll put the link to all of those in the show notes, including Lada's LinkedIn page. So be sure to reach out and connect. Sharing and asking and sharing what we learn, that we grow as a community as well. Lara, it's been a pleasure to have you on the show. Thank you very much for your generosity with your time and your knowledge.
- [36:45]
B
Thank you very much for having me. That was wonderful.
- [36:52]
A
All right, I hope you liked this episode, but before you hit next episode, here's the deal. This podcast is powered by people like you, the members who wanted more than just inspiration. They wanted real tools and real connection to people who are practicing Agile. Every day we're talking access to over 700 hours of agile gold, CTO level strategy talks, summit keynotes, live workshops, E courses, Deep dive interviews, books. And if you're into no estimates, we got the pioneers of no estimates in those deep dive interviews as well. Agile Business intelligence, creating product visions, coaching your product owner courses, you name it. You'll get invites to monthly live Q&As with agile pioneers and practitioners, plus a private Slack community which is free of all of that AI slop you see everywhere. And of course, without the flame wars, it's a community of practitioners that want to learn and thrive together. It's the best place to connect with community and learn together. So if this podcast has helped you before, imagine what you will get from this podcast membership. So head on over to scrummastertoolbox.org membership and join the community that's shaping the future of Agile. We have so much for you, so check out all the details at Scrum Master Toolbox because listening is great, it's important. But doing it together, that's next level. I'll see you in the community. Slack we really hope you liked our show. And if you did, why not rate this podcast on Stitcher or itunes? Share this podcast and let other Scrum Masters know about this valuable resource for the work. Remember that sharing is caring.