Webit 2026: From AI tools to AI teammates: how organisations unlock real impact - Sven Peters

This talk explores the transition of organizations from viewing AI as a mere tool to integrating it as a collaborative teammate in the workplace. The speaker, Sven Peters, elaborates on the evolving role of AI in enhancing productivity and decision-making within teams. He presents findings from a survey of knowledge workers, highlighting that those who adopt AI at the team level report significantly greater productivity gains. Furthermore, the talk emphasizes the importance of creating an AI-ready culture by providing proper context to AI, measuring its impact, and fundamentally rethinking existing workflows. Peters also discusses key strategies for building AI-native teams, including the establishment of knowledge streams and a focused approach to transforming workflows.

Ladies and gentlemen, building an AI-ready company is only the beginning. The next question is even more important. What happens when AI becomes part of the team? For years, organizations have adopted software tools to help people work better. Today, we are entering a new era, an era in which AI is no longer simply a tool we use. It is becoming a collaborator in how work gets done, how

decisions are made, how knowledge is shared, and how teams create value. Our next speaker explores what happens when organizations move beyond AI experimentation [music] and begin redesigning work itself around human-machine collaboration. Please welcome Sven Peters, AI evangelist at Atlassian, with his keynote from AI tools to AI teammates, how organizations unlock real impact. >> Hey, welcome. Hey, wow. Uh, great crowd here. Um, my name is Sven. I'm

an Atlassian's AI evangelist, and I really appreciate that you came here to learn from a real human, right? And not just like ask OpenAI or ChatGPT to just like, how can we how can we, uh, survive in this world of AI? Because, you know, I am actually, can I get the next slide, please? I am actually human. Um, I'm a I'm a human, um, and I, uh,

uh, love to work with with with AI. Um, I have lived experience. I I I I love love AI, but we need to figure out how we work together, right? Humans setting the strategy, making the decisions, but AI actually moving forward. Now, we hear a lot of buzz around AI these days, right? Um, you see, OpenAI says they went from 20 to to 80% of writing code

by just AI. Wow, that's that's impressive, right? Or Anthropic even more, like they say in 3 to 6 months we we will able to 90% of the code will be written by AI. Now, that that is actually there is a lot of buzz from from there because Anthropic CEO said that actually over a year ago and I have not met a company that that says 90% of

the code is written by AI, no enterprise company yet. But the truth is actually all of this will be added to our workflows, right? So, we use AI actually already every day to just like move forward. But the thing is also we are building those AI native teams, these teams that really work with AI and we have to continue to build those. Um, when we look at

how AI got applied, um, we see that and I I'm just like talking to a lot of people, um, that we gain a lot of hours saved, right? You just like write an email with AI, right? You save a lot of hours. You just like summarize this page, all of these things, right? This is really great. So, we did a report, we did a survey with 12,000

knowledge workers and wanted to find out where are we in the state of AI. Now, what people tell us is actually they actually feel like 33% more productive. Wow, that's a big number, right? By just applying a technology, 33% more productivity, that's great. But you also see, um, and I also the leaders are actually saying actually that works got speeded up, right? So, we we get faster

and faster and faster, um, by using actually AI. Now, these are all individual gains. Now, the question is that we next ask like, do you see a return of investment? We asked these executives around the world. And the thing is the numbers are actually not that great. Um, 6% of organizations see a return of investment by using AI. This is not really great, right? So, um, but

said that is I think companies are at different stages of their AI usage. So, some are actually right at the beginning. Some are actually already building agentic workflows. Now, we looked at the 6% and looked at what that are they actually doing differently. And the thing is those companies are actually applying AI at the team level. We call this team multiplier. So, they it apply AI at

the team level and also at the organizational level. So, one level up or two levels up actually. And just 14% of organizations are actually doing this. Now, what is the gain actually? The gain is actually this what this 14% they are 9.4 times more likely to increase And then also they are actually um a lot of more likely. Can I get the next Great. Um, 5.5 times

more likely to have better collaboration and also 13 times more connected actually. So, we see this was really the problem actually before AI came in that to make teams productive. We know that is actually the problem in organizations. So, why not apply AI to that? And we see the numbers are are actually there. So, 14% of the organizations are building those AI native teams. How do they

do that? We also looked into this, right? And they actually doing three things. They actually provide the right context to AI. They are measuring the impact and building AI human workflows. And also they built a AI ready culture. Now, I will dive into all these three things to show you what we mean by that. So, let's talk about context. Context really matters and we heard this from

the previous speaker already, right? 69% of workers say their data isn't ready for AI. So, we need to connect those data points. And as you heard also before, right, these models are super intelligent and they get better and better every day. It's hard really to follow what's the latest model and what's the benefit of the latest model and which model is the best for for for what,

right? So, these models are really But they don't know anything about your company. Where are your decisions? What are your priorities? How do you work, right? Which teams are actually working on what? So, you need to give AI the right context for this. And also you need to help humans to provide AI with the right context. So, how do I speak with AI? How do I do

this, right? All of this is actually necessary. Now, let's talk about what that means, context. So, imagine you want to build a feature. So, you have a work item. If you know who's working on the work item, where did this idea come from, where's the page with all this information, who worked on this page. Um, where are those meeting recordings with decisions that you just like talked

about, where are the documents with with the decisions. Um, where's the design for this feature and so on and so on, right? You can build a real knowledge graph out of your data, right? And at the end you just like know which code actually contributes to that feature. So, we call this actually the teamwork graph and if you use Atlassian tools, you already have access to your

teamwork graph and it will help you just like to find things more. So, second thing is creating impact, right? So, creating impact, you actually need to know your knowledge streams, your value streams, right? Where do you create value? Now, this is a very simple value streams. We say we can go from customer insights to writing those user stories and then we actually writing the code. Very simple,

I know, right? But if your problems is actually customer insights, you can use AI to just like write your user stories automatically or write your code automatically, right? Great. But, your bottleneck will still be gathering those customer insights. You don't create more value at the end if you don't look at your full value scripts. Giving you one example what we do at Atlassian. So, we are actually

gathering customer insights by interviewing customers, right? Through Zoom calls or other other video video technology calls. And then, what we what we going to do Can I get the next slide, please? Thank you. Um what what what what AI is actually doing is actually transcribing those customer calls, right? We use Zoom AI note taker. And then, we have another agent that actually looks at the transcript and

applies a template to this. So, we use that template for years. It's called observations, problems, and opportunities. And I will show you this in a sec. With this template, this AI creates a simple page that a human looks over, right? Is this like hallucination is still a thing. Um to just like make sure it doesn't hallucinate hallucinate. And then, also, what we do, we sending a message

via Slack into a Slack channel, so everyone knows actually what this customer customer interview was all about, right? Great. So, this is the framework that we use. As I said, observations, problems, opportunities. And this is the page that we then create. So, a simple page that we that we create out of AI-generated page, we link the video, so people can dive deeper into this. And then, also,

we send a message, as you see, through a Slack channel. So, all of this is actually available for everyone. They can dive into the data. And our product managers increase actually the customer interviews by 10 times. So, 10 times more customer interviews, so we gather really more insights by the use of AI. Now, said that is 10 times more customer interviews means actually 10 times more pages,

and 10 times more AI-generated pages. And who reads through all those pages? No one, right? >> [laughter] >> It's AI-generated pages. So, this is a problem that we discovered. But, this is not the end, right? We also have another agent that runs every week to find similarities in these interviews. So, we actually identify the real problems that not just one customer has, but multiple customer has. And

we're creating a Confluence whiteboard out of this that the team then uses in their weekly meetings to discuss these findings. So, it actually looks like this. So, the Insights Analyzer agents analyzes the themes, the commonalities that we find in those interviews, puts it on a white whiteboard because we collaborate better on whiteboards. So, the team can actually later go in when they have their their weekly meeting

and discuss those things that they actually found in those customer interviews. Now, this doesn't only mean that we actually drive 10 times more customer interviews. It also means that we actually in those in those meetings have more focused discussions. So, we can actually concentrate more on things and not just anecdotally, I heard this from this one customer that screams the loudest. And that actually means we are

trying to build the right things and really really really increase customer satisfaction with with that. Now, this means we're not getting faster, per se, right? But, we're creating real customer impact. And this is really what we should concentrate on by building those AI building building stuff with AI. All right. and we need to think about our workflows. So, just like throw away throw out your workflows that

you know, and don't just like screw AI into your workflows, but really rethink the way you work. And this is hard work. It's not easy, right? This is really the hard work that you that you need to do. All right, going to the last thing and this is culture. Now, I joined Atlassian for from a long time ago in 2011 and now we have two child 2026.

But I think like things have haven't changed that much. Back in the days, I know I don't know if you know, we have a tool called Jira and use it for Agile software development. Back in the days, I had a lot of discussions around Jira, right? People bought Jira, they certified their Scrum Masters and then they ran sprints and says, "Now we're Agile, but we don't see

the results." So, obviously because this is actually not Agile, right? Agile is mindset, it's a change of working, it's not just a tool and some trainings. And I see exactly the same same things now with AI. Now, companies buy AI tools, they spend time on trainings, and they did just some experimentation or hackathons. Now, and then they call it now AI transformation is done. This is not

the transformation. You need to build a culture around AI. So, don't get me wrong, I think this is important, right? This bottoms-up approach. We need to learn how to deal with this technology. What is working, what is not working. But this is just this bottoms-up approach. You don't just like throw AI to your workers and say, "Have fun with it. By the way, 10% less employees now

because you got AI." No, you just like have to also go a top-down approach, right? So, we set ourselves goals. You see just like we we set ourselves the goals to increase developer joy with AI and change our ways of working by AI adoption. For this actually, for the last thing, we identified 14 hero use cases that our our leadership got measured against, right? The success of

those use cases. And these use cases are not just engineering use cases, these are all over the business. So, we have sales use cases, we have product use cases, HR, legal use cases, marketing use cases in those in those use cases. And then we measure the impact. The impact that we have on AI. So, these are just like examples of agents that we built or changed workflows.

And I would just like to highlight the call prep agents for our sales team that the sales teams use to prep for for customer calls. And it's not the hours that we saved, right? It just makes us faster. No. It is actually what what we what we see is we have 20 to 22% faster deals because we are better prepared for the sales calls. Um and this

is the real impact we want to generate with these hero use cases. But don't get me wrong, right? Um I think trying and learning this technology for your workers is important. Adoption is important, so constantly use AI. But this just speeds you up. This just like makes you faster in the end. Which is okay. This is how you should start. But you also need to measure and

see the impact, know your value streams, and apply AI to your bottlenecks, and also transform your your workflows, your AI human workflows. Transform into this. And this really generates impact on the business at the end. context, impact, and culture are super important things for your organization. So, you need to you need to think about that, how you drive context, impact, and culture to build those AI native

teams. You can dive deeper into the whole data set. The full report is online. Just scan the QR code. It's for free. Um so, we have much more data in the state of teams report, so you can actually see what how you how your organization is is is on this AI journey compared to other organizations. And yeah, thank you very much. Thanks for coming to the talk

and have a great Web IT. Thank you.