Evo Nordics #716 - From Curiosity To Capability - How Organisations Actually Adopt AI

Speaker A: Foreign.

Speaker B: Welcome to the Evolution Exchange podcast. A melting pot of ideas and inspiration shared by some of the most successful technical leaders in the world. The views expressed by the speakers on this podcast are their own and not necessarily representative of their organization.

Speaker C: Welcome to the Evolution Exchange Nordics podcast. We're bringing together the best technical leaders to talk around industry passions and challenges they are facing. I am Louis Wright from Evolution Recruitment Solutions and I help businesses connect with top tech freelance talent. And today I am your host. Today I am joined by Alan Smith from Volvo Group, Sophia Magnuson from Brightnest, Pernilal from Ericsson, and Eli Shalkers from Vattenfall. The views expressed by guest are ah, their own and do not necessarily reflect official position or policy of their organization. And today we are here to discuss the topic from curiosity to capability, how organizations actually adopt AI. So I'd like to hand over to Alan first to introduce himself.

Speaker D: All right, thank you. Nice to be here. So I'm Alan Smith, I work in design and product leadership at Volvo Group and I'm focused on complex internal enterprise products and how we integrate AI into new human agent interactions.

Speaker C: Thanks. Alan and Sophia, if you could introduce yourself.

Speaker A: Yeah, sure. My name is Sophia and I work as ah, head of AI at Brightnest, uh, which is a agency with uh, where we are located in uh, Lulu, Jerelo and Gothenburg. And I focus both on the uh, internal transformation, uh, focus on AI and also the external how to help our customers to um, be able to develop in the best way possible with the help of the AI tools.

Speaker C: Thank you. And Eli, if you could introduce yourself please.

Speaker E: Happy to. Pleased to be here. I'm um, Edleij. I work for Vattenfall. Vattenfall is Sweden's largest producer of electricity. Um, and my role in the company is twofold. On one hand I work for the innovation department and on the other hand I lead a team of data scientists trying to implement AI solutions.

Speaker C: Thank you. And ah, last but not least, Perna, if you could introduce yourself.

Speaker F: Yeah, sure. So. So I'm Prerna, I work at Ericsson and currently I'm part of the digital Transformation team. And we also work in terms of AI adoption within the organization. So currently we experiment with different AI use cases and uh, building AI solutions for the internal team.

Speaker C: Perfect, thank you. And Pranu, if you could start by introducing your question to panel.

Speaker F: So basically uh, when we talk about AI adoption, the first thing that comes to mind is how do you have AI in the organization without the leadership? So I think the leadership or the people at the top level has the very key or important role in AI adoption. So starting with the question, I would say, like, how do you teach the teams in terms of thinking from an AI first mindset instead of them, uh, changing the existing process using AI, I would say like how you can transform the ways of working through AI when it comes to adoption. So could be like building a baseline literacy for the team or building use cases within specific teams. So yeah, the idea is how AI can, instead of adding to the old ways of working, we can have an AI first mindset from the beginning itself.

Speaker E: I think that in order to get an AI mindset, uh, or an AI first mindset, what you really have to do is have a two pronged approach where on one hand you have the people that work at the lowest level of the organization that know the processes that they work with on a daily basis inside out and that they can look at, okay, how can we change these already existing processes? And I know that's not your question, but how can we improve on those? And on the other hand you have the management layer, uh, who has an overview of all the different systems and how they connect to each other. Um, and they are more focusing on the bigger picture. And they see when I look at Fattenfall or my team looks at Fotfo, all we see is code snippets, uh, and that has absolutely nothing to do with electricity. And you want perhaps someone at a management layer who really has a lot of uh, experience to connect all the dots and turn that into something that's relevant for actual energy production. And I think that if you combine these bottom up approach, uh, where you invite people to innovate and to share their thoughts on how we can improve what we already have and combine that with the management experience of everything and how it works together, uh, and the real business cases, that's when you start getting an AI first approach.

Speaker C: Uh, perfect. Thanks Eli and Sophia. Alan, have you any thoughts on that?

Speaker A: Well, I guess the hard question is that how do you do that? You uh, want to align everyone on the same path. And I mean, um, so our company, we're around uh, 45 people. Uh, and it's a totally different thing to work with the transformation with four to five people compared to Vatten Function. So I don't know how many employees are you at bottom pod today?

Speaker E: 20, 20, 1000?

Speaker A: It's like so many questions. How do you do it with that amount of people?

Speaker E: That's a great question. Uh, and I think that even though AI is very important, it's also important to realize that you have your daily operations as well. Uh, you have a lot of people that actually should not be involved in the AI transformation. Uh, I think that it's really important to highlight the key teams, uh, the teams that have the right foundations in place with the right data quality, with the right AI literacy in place already, perhaps through their own projects or through their job, and start there. And I think if you can start there, provide a good use case and then also communicate the use case internally and say like, hey guys, this is what we had. Uh, this was the playing field that we were in and this is what we did and this is the success that it brought. And you can really show from beginning to end this is the impact that it has on our overall business case as a company, not just for one specific team, like we work more efficiently, I can read my emails faster, but really on the business as a whole, then you create this interest in the rest of the organization that kind of ripples out, uh, to inspire, that ripples out to invite people to become more involved in setting up AI systems. And in my experience when you ask someone, please innovate, please do something with AI, they completely stop thinking at all and they cannot come up with anything. Whereas if you say, hey, we have used this specific system to solve this specific problem, then they are already in this flow of uh, creativity where they can say, well, okay, if we can use this system for that approach, we can repurpose that for something that we are working on. So I really think that it's a seed that needs to grow and I think that uh, starting with the most, or a low hanging fruit, I know that that's been said a thousand times, uh, starting there, letting it ripple out. That is the best approach for me to get an AI first mindset.

Speaker F: I think there's a lot of change management in the whole process because you're not just changing like adding AI in your organization, you're changing the ways of working. And people are very used to the traditional old ways of working. And when you bring AI into the picture, it's basically redesigning the whole workflow altogether. It's not just doing the same tasks with AI, it's basically how you can work efficiently overall. So you have to change the mindset of people as well, I feel, because there's always this constraint when it comes to AI, especially with the people who are not very literate in terms of using the AI tools. So they're always, uh, finding ways to criticize that why AI won't work in our use case or in our scenario. I mean that's my experience so far.

Speaker D: Something I found it's very practical. But uh, if you work in a big company or medium sized company, here's something I've discovered that is pretty high signal. Look for Excel sheets, and in particular look for really complicated Excel sheets. You've probably all seen some version of this where someone shows you their Excel sheet. It takes like three minutes to load. Uh, it's practically an application in and of itself. That's usually a signal that, that is ready to be turned into a process that can be either automated by AI or at least assisted, uh, and keep the human in the loop. So, um, you can imagine, uh, that Excel sheet essentially models some sort of internal workflow and then map where that person who works with that sheet or those people, where are they searching for other people to uh, hand off outputs or get inputs? How are they synthesizing, um, and reformatting data that comes out of that Excel sheet and then work with that team to ask how AI might be able to shape that change of work. How would reasoning and automation be able, uh, to be applicable in that situation?

Speaker E: Uh, and I think you bring up a good point there. Not every team is ready to adopt AI right now, even though sometimes, like goose liver, we want to shove it down their throats. And I think it's important to invite teams to think along and not say you shall now use AI, but to really say, hey guys, we see that you have this complicated Excel file. We see that you have this process that potentially could be improved. And that way you're also showing them the value of uh, to what is relevant to them rather than just saying thou shalt use.

Speaker D: Yeah, and actually, and what I've learned is actually those folks, uh, are typically, they may not care about AI so much, but what they care about is that you're making their job less stressful and you're making them more effective and generally just better at whatever they're doing.

Speaker C: Great insights to start off with the question. And Alan, if you'd like to introduce your question to the panel.

Speaker D: Sure. So, um, yeah, it's very highly related to this one and it's really thrown it out there to this group is whether organizations actually are ready for AI agents or are most of us still struggling with the basics. Um, fragmented processes, unclear ownership, poor data quality, legacy systems that were clearly never designed for autonomous workflows. This is very pervasive in um, big enterprises that I'm interacting with. And I'm curious For these folks, what they think might need to change before agents can become more than demos.

Speaker A: I take a guess here and say that I think it's a big difference if you have a smaller or a big company. So with us as well, we're pretty big but not compared to you, um, but with like four to five people, um, and doing that uh, transportation formation all over the company at the moment I have the possibility to actually get everyone to learn the basics in a very quick way but in order, uh, when they do that they will also understand the whole agent structure. Uh and I can already see that it goes quick and that in our case I'd say that uh, we are mature for that. But I'm not sure it is also because I'm so involved and ah, I have the management with me and uh, we're working with like establishing everything with mindset, ambition etc from the both bottom and like um, top down, um, but also speaking with everyone. So it feels like we have the possibility to have the discussions and understand that everyone understands. But as a bigger company, uh, I would guess that it's a totally different ballgame.

Speaker E: At least for us it is. Uh, I think that there's no such thing in FATUFA as organizational readiness for AI, but I do think that we have team and department readiness for AI because you have an extreme heterogeneity uh, in terms of the people that we employ. We have people that are office workers, we have people that are field workers, we have people that have been in the same role for the past past 500 years and we have been people that started yesterday. Um, and all of these people have such a variety in their AI literacy, their understanding of the corporate uh, systems as well. And I think that it's really important or that you build organizational readiness by doing this one team at a time. Uh, you have islands within the organization that are very mature. Uh, let's say for example the HR process, the HR was there since day one of Fattenfall, um, and they have very well established systems, they have very well documented uh, databases, they very well everything and that is a process that way you can assist with automation. Whereas if you look at a very new department, so at the moment we're starting to work on new nuclear energy uh within Sweden. That is a department where they're still struggling with uh, the regulations that are ever changing by the Swedish government and they don't have their processes yet in place. So I think that it's important to address the question to the right part of the organization. The More stable they are, the better their foundation is, uh, the more easy they will have a time of adopting AI.

Speaker F: I agree in the sense that in a bigger organization we have different heterogeneous groups of people and they come from a different skill set altogether. Some have a good AI literacy, some don't. But the biggest challenge I see in large organizations or enterprises is the data, because people have very scattered or unstructured data. Of course you can build AI models on unstructured data, but for that also you need to do processing and a lot of other things which not everyone has access to. So even for like API calls or for the MCP servers, you need different accesses to different tools to connect. For example, if you're building an agent to connect say two to three different systems together in this organization, maybe just say for a smaller team you need access to those different tools and then you need a whole process to get that access. So I feel that it's possible to have agents in a smaller team within a larger organization. But to move from a uh, demo or like uh, I would say like a test agent to agent which is usable across the organization, you need a lot of accesses, you need a lot of approvals and that process is a very constraining process and a lot of people just give up in that. So we feel that people are building agents in the organization, but it's usually at a very small scale, say for a smaller team or a smaller organization, but organization wide. If you talk about AI agents, I think that's still a dream in my experience.

Speaker C: Mhm.

Speaker D: A lot of security concerns also in it around that as well. Yeah, it seems like it's gonna uh, there's a cohort of companies there may be more traditional, older and it's gonna be adoption that's very bottoms up, maybe to some degree and experimental. And over the next five years, 10 years, maybe we'll see bigger implementations uh, where it's clearly been de risked and there's a very clear ROI on it. And then you have this other class of companies that are tech companies and I think are already adopting this um, quite rapidly. And uh, for me that's where I see the split. I guess it would have been interesting to maybe a more tech oriented company uh, represented here. But um, that's at least what I'm seeing.

Speaker E: I had a question for you Alan, uh, to follow up on that. Do you think that there's a difference between companies that are listed on the stock exchange and companies that are not listed on the stock exchange?

Speaker C: No.

Speaker D: And the reason I say that is uh, there was a article a few weeks ago, someone at Amazon and they had uh, they were doing automated deployments of code and from an agent and it wiped out, it wiped out some big chunk of their code base and people were not able to even access Amazon. Now think about that. The, the largest storefront on earth. And so they reverted back to previous good uh, state and then they put in a process where all check uh ins have to be human reviewed now by, by a senior engineer. Um, they're very public. They're one of the biggest companies in the world. I think it's uh. And we also see lots of other examples of publicly listed companies like Meta, uh, Google, whatever, who are very open about laying off lots of people and in part replacing a lot of those roles, uh, with AI and agents.

Speaker E: But you have two separate topics here and uh, hooking onto your first one. Uh, I think that database wipes and stuff like that can be easily avoided by just following the already existing IT standards. I'm stationed in IT and we have a different development environment and a uh, testing environment and acceptance environment and production environment environment. And indeed the people that have access to make any code change at all in the production environment should already be the senior engineers. So I wouldn't want to put the blame on this, uh, for this on AI, but rather just poor IT management

Speaker D: coming to the defense of AI. Good.

Speaker E: But do you think that people are actually getting laid off, um, in the European markets as well in companies such as Volvo? Is this something that you see around you?

Speaker D: Uh, I mean, nothing. Um, I don't see it as much really. I don't know, maybe I'm not clued in as much on the European market, but you definitely don't hear about it as much. And so either. That's because in Europe there's a different stance on it, or maybe the adoption is slower here. Um, it's a different regulatory environment. It could be different reasons, I'm not sure.

Speaker E: Or perhaps the 10x that the big tech companies promise us, 10x or 10 times improvement of the amount of work that you can do is simply not true. Because I think that AI definitely speeds up a lot of processes. Um, for example, co generation is super fast right now, but you still get stuck in other parts of the process which is requirements gathering, uh, which is stakeholder engagement, which is getting the approval that Prena was talking about, um, making sure that all the systems are in place and uh, struggling with IT tickets. And I think that yes, we can definitely speed up Part of the workflow and that part might be sped up 10x. But I think that the entire project lifespan has not drastically decreased. Would you agree with that, Sofia?

Speaker A: Uh, I think it's interesting. I was just talking to someone else about this. Uh, I think it was yesterday, like, will AI take your job or not? And how does it really look in Europe and in Sweden at the moment? And I think it's interesting to keep the eyes on us and what's going on there and because, uh, as you say, Alan, um, I agree that I haven't seen that much layoffs with like. The explanation is that it's because of AI. Uh, so far, however, I have heard of some, um, But I'm also thinking that this is a phase. So I am expecting companies to maybe do this, but then again to also rearrange their, uh, internal processes and find new roles and um, employ people for different tasks. Uh, in. I don't know. We don't know that much about the future at the moment, but I don't know. In a year, two, three, I don't know. So that's why I'm thinking, that's why I think it's interesting to look at the US and to see what's happening there. Now we see a lot of layoffs, but in like a year we really see that the same people are and project to those companies, but with um, different tasks. Because now we have. The companies have started to do X, Y and Z. Um, so I think it's an ah. I think it's very. It's an interesting time that we're living in.

Speaker C: And Sophia, if nobody's got anything else to add on that, would you like to introduce your question as well?

Speaker A: Yeah, sure. And that is, uh, also kind of the, the same topic, uh, because from what I'm seeing, a lot of AI adoption ends up being about speeding up existing processes rather than questioning whether those processes should look the same at all. So I would very much like to hear what your experience is on this and if you've seen organizations that actually make that shift and what did it take?

Speaker D: This is such a good question, uh,

Speaker C: because

Speaker D: the answer is almost inferred in the question, which is to really get the most out of this new capability, you do have to refactor the processes, you have to delete stuff, you have to rethink. Um, the pattern I'm seeing, uh, in general across big companies is we're sprinkling some AI seasoning on all the existing stuff, trying to optimize those specific tasks. But then we're sub optimizing whole. And so um, one of the things I'm working with quite closely is how you, how you work with an end to end workflow, a complex workflow and be able to go through that and refactor it completely, not with just AI I would say that can be a big part of it. But there's processes that have existed for 20, 30 years in some large enterprises where uh, there's still a fax machine involved or there's uh, an interaction with a COBOL mainframe that's involved. And so um, I think uh, there's this big opportunity now maybe because AI has such energy to really motivate Reef just going through and redesigning these workflows in a totally new way. Whether it's with AI or not. A lot can be done potentially um, without it. But I think with it um, for certainly it'll set companies up for success.

Speaker E: But at the end of the day it all boils down to a business case, right? And somebody needs to pay for this transformation. And I think that it's very important that when you want to change the workflows that you're working on is that you start by identifying hey, what could be the potential value if we swap out the fax machine? What could be the potential value you if we swap out the cobalt mainframe and compare that to the cost of what it would take to implement AI. And I think that sometimes just because we can do things more efficiently doesn't mean that we have to uh, sometimes a winning concept is good enough. And I think that it's important to prioritize these different use cases and say like hey, we really want to focus on the AI use cases that really touch the core of our business or that have a uh, ripple down effect through the entire organization or that really uh, affect um, those that are really fundamental, really valuable. Because if you're going to do an AI project for swapping out a fax machine, probably the fax machine is still there because this workflow gets used three times a year. Um, so yeah, and if you spend money there on improving that, maybe no one will notice the result and maybe you will be wasting in quotation marks, uh, money in improving something that doesn't really matter. I think it's really important to identify the right thing first and then try to start with the demo and really try to follow up on it to make sure that it ends up in production. And the difference between those two is the quality of the business case.

Speaker D: This is where good user research Comes in, um, because part of that business case is going in and talking to folks and watching them do their jobs in order to get a really good qualitative feel in addition to any quantitative data you have around this,

Speaker C: to

Speaker D: see the full 360 of what that business case is. And um, because it is a business case, it's also a uh, let's say employee well being case. A lot of these uh, point solutions that have just stacked up over 20, 30 years makes it really complicated, unnecessarily complicated for people to do their jobs. And we see that results in a lot of burnout, results in a lot of mistakes and stress on people.

Speaker F: I think it's also the way AI is introduced within the teams. So for example, if you're working in a transformation project and you want to automate certain processes using AI, the users usually have this mindset that it's a magic wand. You know, it's it AI in a way translates to optimization, but that's not actually true always. I mean of course AI can lead to optimization, but it's also like how you deploy it. So I feel that communication and understanding of what AI can do, as you said, building use cases and telling people that this is where AI sits in, this is the impact you can have given you fulfillment. Like these 5 to 10 criteria or requirements for this to work efficiently. So I would say like when you communicate to your stakeholders, you should be more open in terms of what could work if this is the best case scenario and what could not work if this is not being followed or this is not being changed. Because otherwise there's no point of bringing AI to the team in a way. Because I feel that a lot of times when you work with customers or stakeholders, they have this expectation that AI is the solution to all our problems and AI will help us fix this situation in two, three months. But that's not actually the case. So I feel that understanding uh, where AI sits in building that use case, identifying which are the key value drivers, which are the key uh, inputs which goes into the use case and the expected outcome that needs to be clearly communicated when you build an AI solution within the organization.

Speaker C: Thanks there Perna. And has anybody else got anything else they'd like to add on to that?

Speaker F: No.

Speaker C: So Eli, if you could introduce your question please, we'd be happy to.

Speaker E: Um, we already have AI agents in basically every organization. Those can be some custom made, they can be made by a bigger organization such as Microsoft. Um, but at some point right now they're all still Human in the loop systems. And I think that the value of AI really starts to come through if at some point we are able to make this automation step, to have a fully autonomous system. We are not there yet. Or I think at least most organizations are not there yet. Ah, so my question to you is what is needed in terms of trust before we can get these more autonomous AI systems?

Speaker D: For me, trust, it's not a um, model performance question, uh, at least in my experience it's a design problem. Uh, so when I'm working with folks, they want to know uh, where did this data come from? Can I trust it? Is there traceability, uh, through this system, is there visibility? Black boxes make people a little nervous, uh, especially when they're dealing with complex supply chains and potentially tens of millions or hundreds of millions of dollars, uh, worth of, of uh, goods. And can they understand the confidence level? Can they override it? Ah, so ah, like all of these things are kind of more design problems um, around how AI is implemented and what that interaction model is between um, the AI model and the human.

Speaker A: Yeah. And I also like to add that when it comes to like getting organizations to trust their agents or use them to really uh, get the leaders uh, involved and to get them to test and work and also show the amount of mistakes they do, but also the amount of success they have using it. Uh, but to really um, like dare to work on that with that test country to get people to just, just to start doing it, to start using them.

Speaker F: I, I feel that there are two aspects to your question. One could be what are the tasks you can fully automate? So for example scheduling a meeting, of course it needs to be fully automated. I don't have to go to teams and do things myself. But when it comes to taking business decisions, I feel it's not just now with the AI. So in my prior experience I used to build machine learning models and people were not comfortable. They need explainability of those ML models and when it comes to AI, it's more black box. So it's not about trusting in my opinion. It's about what is at stake, what is the risk that they carry if things don't work, if the model hallucinates and say maybe you're having a stakeholder engagement through an AI agent, but what if the agent goes and tells the incorrect information? So then what's the cost of that? That's something that we need to evaluate when we talk about fully automation.

Speaker E: Fair points. Thank you all for sharing your thoughts. I would like to ask perhaps a Follow up question. Because there are so many uh, different AI initiatives at least in our organization, I guess, guessing your organizations as well. How do you choose which one to prioritize? How do you choose which one to give a bucket of money and which one not to give it to? Seems like it's an easy question.

Speaker F: I would say the prioritization in terms of AI depends on firstly your readiness, how ready you are in terms of building the solution, access to data as you mentioned, access to these IT related things where you need like access to different API calls and all. Second could be your value which you're delivering the customers or the users that you are impacting within the organization. So is it a high impact use case or a low impact feasibility? How feasible? As I mentioned in the previous point, then I would say like will it give return um, in the immediately or do you think it's a long investment project? So I think there could be like factors.

Speaker D: Yeah, I like that. That was uh, similar to how I was thinking of it. There's uh, you know, there's some sort of two by two right where it's impact and then what's the risk that things blow up and uh, if you can find something that's high impact and you know, low risk that it's going to go south on you, those are good ones. Uh, potentially even the low impact wins, uh where the cost of failure isn't high are also worth doing if the investment is proportional to it. Just to you know, going back to what others have said earlier in a way we need to build this new muscle and so the more reps we can get in, in different places in the organization, uh, the better. And if we can give people opportunities to sort of build and do uh, real work things utilizing this new capability, uh, where it's low risk and they don't have to worry about getting into trouble if uh, it doesn't work quite right the first time. Um, then those could potentially be pretty good opportunities as well.

Speaker C: Yeah.

Speaker E: So on one hand you say Alan, that we should start with the smaller stuff. Uh, whereas Prerna is saying we should maybe focus on the bigger stuff where you can actually have an impact. And I think that there's something to be said for both. Uh, and perhaps you need a mix in your portfolio. I mean you're never going to prioritize just a single AI use case. You'll be prioritizing some low hanging fruit and some m. Moonshots.

Speaker A: And I can just add that from our perspective as a consult consultancy agency, we uh, we listen to our customers. So if they come to us or, uh, with a specific problem, of course we solve it for them. So that is kind of how our prioritization is, is steered. And then we do a lot of stuff internally, but as we're where we have our tech team and that is the prioritization, uh, for them to learn and build and to just, uh, move forward and to get those solutions out in the company. But I'd say that primarily, uh, it's our clients that, uh, is, um, the biggest factor, of course, and you have

Speaker E: other companies as clients. But I think that the same applies to Volvo, Ericsson and Vattenfadel, where we should be listening to our customers as well and trying to deliver what they need. And if that's perhaps, um, more accurate insights into how you're spending your energy and your electricity, uh, then that's where we should build it. So, yeah, I think it makes sense to follow customer demand.

Speaker C: Is there any topics anybody, or question, anything that anybody would like to revisit?

Speaker F: One question. So when we talk about AI adoption, you talk about AI literacy. And when you talk about AI literacy, we have new tools coming up every few days. So when you tell your employees or colleagues that, okay, you should, you know, move m towards learning AI and, uh, upgrading your skills. But then should you be, like, expanding in depth because maybe, like, learn a tool really well, or should you be expanding under breadth, like learn Claude how to use different tools and be like, I would say, like, knowledgeable in different tools instead of being a master of one tool. What's your suggestion?

Speaker E: I think that no enterprise system is built by a single person. I think that you will automatically be looking at a team. And in a team you'll need both. You will need some people that really can connect the dots. And perhaps an architect can say, like, oh, we have Claude, we have GitHub Copilot. Which of those two is best for solving this problem and then handing it over to the corresponding team, um, to actually do the implementation? So I would like to separate the design and execution phase. In the design phase, you need broad knowledge, and in the execution, deep knowledge.

Speaker D: That's a good way of putting it. I think this, this is, again, me speaking personally. Uh, but I think the generic AI training really has limited value unless people can apply it to their actual work. And there are early adopters. I would put myself in that camp where I'm all in and I'm always trying out new tools. I'm, uh, going to different GitHub repos running all kinds of different software. Uh, that's a small group of people. Most, the majority of people, I don't think they really care about tools and what's effective for them.

Speaker A: Is

Speaker D: one, are they curious, do they have sort of the right mindset and are they open and things like that. Um, but for the really effective learning it's taking a real workflow and then redesigning it with them, uh, with AI in the loop. And this is, this is kind of goes back to that Excel sheet signal I uh, was pointing to because that's where I've had some good um, success.

Speaker A: And I can just add that my focus is to get everyone kind of up to speed and aligned and just to kind of make sure that they are on the train. So ah, we are focusing on um, flow down for example just to get everyone up and running and not to, just to make it easier for them to ask each other and to test stuff and to discuss. And then of course as we have design we have a lot of different skills. Of course it's going to be different tools that they're going to use as well. Both uh, for example, uh, code and codex for example. So uh, but just to get everyone up and running, uh, to really make it easy for them.

Speaker F: No, I definitely agree and I agree to the point where you mentioned is that, that Sorry Ellie. That this broad literacy is good for people who are in a way like working with AI but not building it. So people who are maybe coming up with use cases, people who are in the business. So it's good to know like basic of every tool which exists in the organization and maybe outside. But it's important for the developers at least to know in depth on how to use that tool. So yeah, I agree.

Speaker C: Anything anyone else would like to add? No, no problem. So before we end the podcast I'd like to say thanks so much to all of our guests for sharing their thoughts in today's conversation. Once again, our uh, guests today on the podcast have been Alan Smith from Volvo Group, Sophia Magnussen from Brightnest, Fernal from Ericsson, and Eli Shalkers from Vattenvall. If you are hiring for new technical or looking for a new role, feel free to get in touch with us at Evolution. Or if anyone you know would like to be featured on a future podcast, you can message me too. I am Louie Wright and you can find me on LinkedIn or visit us@ovolutionjobs.com Nordics thanks again to all of our guests and thank you for listening. We hope you can join us next time.