3248: Atlassian - When AI Becomes a Teammate, Not Just a Tool

[00:00:04] How do you lead people through a technological revolution, one that is rewriting how we work, communicate and create, and do so in a way that ensures that nobody gets left behind? Well, I recently bumped into Avani. She's the Chief People Officer at Atlassian. I bumped into her on the show floor at the Team 25 event.

[00:00:26] And we sat down to talk about how, as AI becomes more embedded in the modern workplace, why there's a growing urgency for leaders to rethink not just the tools that we're using, but the ways that we structure our teams, the way we communicate how that needs to change, and how employers can support continuous learning.

[00:00:47] So we discussed how Atlassian is approaching this evolution from their side, from transforming onboarding with homegrown agents to replacing traditional meetings with asynchronous tools like Loom. Moving away from email completely, that's something you're going to want to hear more about, and how their distributed first approach is building a more inclusive, people-centered way of working.

[00:01:11] So whether you're wondering how you're going to successfully upskill your teams to make them ready for an AI-powered future, or grappling with some of the practicalities of breaking down silos in a global workforce, this conversation will offer a candid look behind the curtain of one of the world's most forward-thinking tech companies. But enough scene setting for me. It's time for me to officially introduce you to today's guest.

[00:01:37] So thank you for joining me here at Team 25 in Anaheim. For everyone listening, can you tell them a little about who you are and what you do? Nice to meet you, Neil. My name is Avni Prabhakar. I'm the Chief People Officer of Atlassian. What I do, apart from leading a people team, is I work very, very closely with the business leaders to really focus on what are the key things and how do we shape up a talent strategy to make the maximum impact on the customer outcomes.

[00:02:04] So I play, I would say, I play a lot in the field with my business leaders at all levels to make sure that they have the, are we supporting them, solving all their talent problems, to make sure that they achieve time to value, not time to market or speed to market, but time to value for our customers. And there's been so many big announcements this year, especially around AI.

[00:02:29] And one of the things that you've spoken about that caught my attention was not about the technology, but the actual mindset shift that's needed to embrace AI as a collaborator. So on that side of things, what steps have you taken within Atlassian's people to support that shift, to make AI feel less intimidating and more empowering for anyone that feels that pace of technological change a little bit overwhelming? So for us, before we give any advice, we want to make sure that it's backed with data.

[00:02:58] So we did our AI collaboration report. And what we saw when we interviewed a lot of knowledge workers across Fortune 500 companies is they are, everyone is at a different journey at this point in time. So if I was to give you like, if I was, there are about five stages, but if I was to bucket them in two stages, there's one bucket of where people are either just playing with it. Okay, fine, chat, GBT will give me an answer to something.

[00:03:22] Then there is a second bucket in there where people are kind of asking a little bit more question to say like, hey, I'm using Copilot in my email. And then can you tell me what are my next meetings and calendars? So I would say those people are just tinkering with it. And then once you move to the second bucket, where you actually start asking AI, I'm a market researcher. And can you tell me in my customer base, how many of them, what is the user behavior? Can you get me some data? So you get to like stage two.

[00:03:52] You get to stage four, and which is what I call strategic collaborator, is when you start sparring with AI as like your teammate. So say like, hey, this is my hypothesis. But if I changed it, how do you think the customer behavior will change? So then that market research person is kind of becoming a true strategic collaborator and using AI as that.

[00:04:13] When you do that, I think what you see is it's a shift in mindset of how you're using it rather than saying that, hey, I just want to learn a new technology and tool. That's when it becomes intimidating. It's almost like, oh, this is a new course I need to take to really get smarter. Right. So people who are using it as a strategic collaborator, they're becoming more and more innovative and creative because it's freeing up a lot amount of time for them.

[00:04:38] And what they are realizing when that is happening is like, hey, it's not that now that I have extra time, my job is going to get redundant. It is more like, hey, now that I have extra time, I'm actually able to do my job at A++ level that I was doing before. So unless that doesn't become your own experience, it is always intimidating. So I'll give an example of larger corporates where AI has become like this. We wrote the digital transformation wave, agile transformation wave. And now it's like this AI transformation.

[00:05:07] So if I'm an employee sitting in a large multinational company, I'm thinking like, oh, my God, now what does it mean? Restructure another reorg and then this will impact my job. So I feel like that will be my advice. Like don't AI is not that way. AI has to be more grounds up where everyone needs to feel like I am an AI native. I can tinker with it and then really experience like how it is empowering you to do your current job better. So that requires a mind shift change. And that happens only when you're on the journey.

[00:05:36] Your question about how we have done it at Atlassian in terms of for us, I would say like we have an unfair advantage because we are also the creator of technology. So we are creating it. So for us, there is a general excitement in the organization. Hey, Robo, let's go. And every employee is using it. Not because they're worried about like, hey, what does it mean? It's like, hey, we are building cool shit. Let's use it ourselves. So I feel like we have a we are in a different board when it comes to that. Love it. So many great points there.

[00:06:04] And another reason I was excited to speak with you today is in your recent session with so many leaders from General Motors, Vanguard and so many others. Disconnection and agility came up as key concerns, especially about working in silos, etc. So how are you at Atlassian? How are you addressing those issues internally? And how can other organizations and business leaders listening? How can they learn from this distributed first approach?

[00:06:28] I would say if you ask any of the Fortune 500 companies who have reached a certain scale, if you ask them, hey, what are your top three kryptonites? They will be around working in silos, perfection or progress, whatever you call flow versus static, this thing. The other one that you will see is the dependency loop. Everyone feels like I cannot move forward until this person moves. So even though they all are siloed, but they still feel dependent on each other. So it's a very interesting thing.

[00:06:56] So and I would say it also depends on as an organization, how I shouldn't say how old you are, but what is your age? So the younger you are in your journey, it's faster for you to adapt newer working practices. The older you are in terms of inheritance that you have, like the big multinationals who are like doing extremely good work, but then they have a lot of other constraints that kind of make their thinking a little bit narrower. So if you are in a financial banking insurance, like that sector, you do have to work under some regulatory requirements.

[00:07:27] And that also limits you. So I also want to be very honest in this answer. Like, hey, I cannot come with like this cool playbook to say, like, do it because this is how a new tech company is doing. You can do it. No, every organization is on a different journey. So you have to. So I want to lay that up front. And my what I can share with you in terms of what we are still doing and what you can do, my advice usually to my customers or peers that I talk to in that environment is like start small. So pick one team.

[00:07:54] I'm sure there'll always be a small tech team and start from a tech driven team because it's they work on different frameworks. Usually a tech team, even in a big multinational bank, would be working on agile. Their tech stack and tooling will look different. And they would. So for them to adapt some of these working practices and become like a living example within a bigger organization is much easier for the change adoption. So my only my first advice to them is like start small.

[00:08:19] Pick a team, not very massive, under 500, 200 odd team and start with ways of working. When I say ways of working, biggest pain point meetings. Just ask the team like, hey, what is hurting the most? I can tell you 80 percent of them will say like, I'm tired. Like it's it's the meeting fatigue. So start with changing the meeting culture. Change it with practices. When I say practices, first start saying, hey, first question we'll ask, do we really need this meeting?

[00:08:46] OK, have a framework to say you ask like, are we making a decision? Is it an informed meeting? Is it consult meeting? Like come up with a framework so that everyone knows like this meeting is we're not making any decision. So people can decide, you know, who needs to be in and out of the meeting. Secondly, replace it with some tooling. So if you're not, if it's just an informed meeting, I'm coming and telling you record it. Use a tool like Loom, which is an asynchronous way. You just record a Loom meeting to say like, hey, all of you working across different time zones. This is the content.

[00:09:16] And then people can react to it in their own time and space. So that kind of saves up a lot of time. It's like modern ways of working. If you ask me, have to move to an asynchronous ways of working. So really push the meetings deep, push the thinking hard to get away from this having to be in front of each other trying to work. So that's one, I would say, one practice I can share from a modern ways of working. Second practice, I would say, would be in that dependency loop. Be very clear about like who is the ultimate decision maker?

[00:09:45] Because at times if you ask people like, are you the decision maker? Am I the decision? And then they're like, oh, we all making decisions. No, no, just make the tough call that saves up a lot of time. It gives a lot of clarity to the team who's making what a decision. And third thing I would say is just set very, very clear. And I know it sounds very archaic. So not clear goals, but I would say just set very clear milestones to get to those goals. In terms of we hit this milestone, we all are aligned.

[00:10:14] So if you do these three things in terms of setting clear goals, who's making the decision and then change your ways of working by just making one trick about meetings would be like my quick soundbite on it. I can go very deep on it because that's where we spend all our time, like how to make teams clicking. But just in the interest of time, I'll probably just say those three things. Well, love it.

[00:10:36] And as AI continues to reshape how every team works, how are you rethinking talent development from hiring to upskilling and ultimately ensure people and technology evolve together and nobody gets left behind? Anything you can share around that? So I would say, I mean, this is what I think day in and day out in terms of how do we really raise the bar on performance and talent in an organization?

[00:11:02] Interestingly, if you ask any of the CEOs right now, the top three things, what are their top priority? Number one priority. I would say all three of them will be in the people category. Starting with like, how do I make sure that my top talent is working on my top 20% of the problem? It sounds good. But when you really look at any organization, the big problems, you don't usually have your top talent working on it because we are so siloed. So I could be good at doing something, but because I'm sitting here, I'm only doing and maybe I'm not doing the most important thing because that's happening somewhere else.

[00:11:31] So the reason I'm giving you the backdrop is that it's not just people or a CPO problem. If you ask any of the CEOs, they will say that's the top talent. Second thing is like, how do I optimize? How do I make sure that even the rest of my talent in the company is really kind of I'm optimizing for their time really well? So, for example, currently people spend like you can be did our survey and they said 76% of the people are not doing the most important work in the company.

[00:11:58] They spend a lot of their time searching information at this point in time, right? So how do we think about it in terms of upskilling and hiring and everything is definitely go deep in terms of your hiring processes and keep evolving them on, I would say, almost on a quarterly basis. So if you have this coding interview or if you have this set of selection criteria, just revisit every quarter to say, are they the most relevant right now based on what we need?

[00:12:25] In order to do that, people always go this route like, hey, it's skill mapping and I need to do the comp. No, no, no. That's where you get too caught up in this very big enterprise. See, this thing, just keep it simple. On a quarter on quarter basis, what are the top priorities as a company you're working on? Really look at the hiring funnel to say like what percentage are we hiring to meet those?

[00:12:46] In that, make sure that you go one click below to look at the selection criteria to make sure do a quarterly market mapping or which companies are doing similar stuff. Because the landscape is changing so fast. You have to stay like it's not like, hey, I know these are the four companies I hire from and then I just go there or this is how you're mapping. No, you have to get really, really agile about how you do that. So that's on the hiring side and you need to have a machine that does that.

[00:13:14] On the upskilling side, usually everyone would say on the AI wave internally, that was the number one thing, like even with our customers, like how do we upskill our employees to kind of start using AI? I don't think so there's any dearth of AI tools in any of the companies right now. Everyone is kind of because of FOMO. So let's get it in because it's also you have to get a tick. Yes, we are driving AI transformation and we have deployed this.

[00:13:39] I think it's the AI adoption is what everyone is kind of working on or learning from. So when I say upskilling that you can do it multiple ways, get tools which helps your non tech users to build. Right now, what happens is you get one tool. I'm just using chat GPD copilot as in this thing. And then everyone feels like I have to go and just keep tinkering with it. That's not where the adoption and the learning happens.

[00:14:05] If you have to really do upskilling, look at something like like we've built like a studio builder, right? But I as a HR person can go and build an agent, right? That's where the excitement and the learning comes from. That's real upskilling on the job. If you're a leader, come up with specific use cases top down and then ask your team to go and solve for them using AI rather than thinking through it from a process lens and transformation because that's a traditional way of thinking. So I'll give you an example.

[00:14:35] Onboarding, right? In HR space, we have been like everyone has been working on onboarding forever. It's a constant. This thing this time we said, hey, you know what? We have robo and we have studio. Let's go and build our own onboarding agent. So we build an onboarding agent called Nora. Our team did it overnight with no support from IT. It was built by people team. So there's a certain excitement when you build something yourself and they created it and they said, OK, fine, let's launch it for our first onboarding hires for this week.

[00:15:05] Day one, they were like 800 queries of me. Like there's just so much excitement. It's built by them. We are tracking it. And what is it? It gets exciting because now they're not just tracking the action. They're tracking the insights coming out of it to say like, hey, 80% of the questions is coming about mobility. They just want to know where this thing is. So I feel like then it's more organic.

[00:15:27] So that's how I'm kind of thinking about hiring and then upskilling and upskilling more in terms of making sure that you use the right tools, which they can build and play with rather than deploying a company-wide this thing to say like, hey, now your copilot is there sitting in your email. You can use it the way you want to. You won't go anywhere. With that, I have lots of use cases. I think internally within people team itself, we have built some 600 custom agents for different use cases. There's a rewards program.

[00:15:57] I have X number of nominations. I want to get to Y. AI will help you get down to that. Our annual performance cycle, we call that as Apex. And we have an Apex agent, which we build for employees. So if I need to sit with you and do my performance review, like I'll just go to the Apex agent and say like, this is what I did the entire year. These are my list of things. Write it down for me or tell me. And they're loving it. So that's how like find the real use cases in the employee life cycle and it becomes more and more. That's how you get traction.

[00:16:27] Yeah. And I love the amount of time that can be saved doing that as well. And knowing I was going to meet you in person today, I always like to do a bit of research on my guests. And one of the things I noticed about you, there was a phrase, people-centered communication that stood out from a recent LinkedIn post of yours. So can you unpack what that looks like in practice, especially when rolling out new AI tools globally across maybe a distributed workforce? What does that people-centered communication look like?

[00:16:54] I would say first, it's definitely driven by the value system that you come from. And I think at last, and being a very value-driven company, with or without AI, I would say it has always been people-centered. I would say in the distributed work environment with AI, we are using that as a very like as an advantage. So I'll give you some examples of if you're doing a lot of asynchronous ways of working, which means most of your communication is written down, right?

[00:17:21] We do have some AI and some tooling like which does sentiment analysis right behind the scene to say like if you can read this blog or this thing, what are you really sensing from this thing? So that gives you even more deeper insights rather than me having to sit with every employee or you're not in a town hall in person to really understand what's happening. So I feel like in distributed work and with the AI, the way the AI is moving, it's becoming even more and more important that you leverage AI to get sentiment.

[00:17:49] And hence that helps you with a more people-centered communication approach. That's one. It also becomes real-time. So like no one is have to wait. Like I feel like the response time needs to also be in the moment. That's when it becomes more and more authentic. So we are using our own chat features to make sure that it's in real time.

[00:18:10] And if you look at all the work that you've done here, what would you say are some of the early lessons Atlassian has learned from experimenting with AI in HR and people ops and everything from onboarding to engagement or even performance feedback? Because I think there'll be many business leaders listening there at the very beginning of that journey. And you're somewhat ahead because you're playing with that cool technology that you've helped build to solve these problems. Any early lessons that you learned along the way that would be helpful?

[00:18:39] So I'll probably try and give you specific examples and the use cases so that it becomes real for the users. So if I shift gears from AI and robot and how we're building AI in our tools, so Jira service management as HR, we are the power users of it. So if any employees right now in a distributed environment, the most thing they struggle with is finding information. Who do I ask? Where do I go? It looks very different. So we have our Good Day employee hub, which is built on our confluence as a company hub.

[00:19:08] So we leverage that. And then Jira service management is our Good Day help desk. So that's where an employee goes in and they ask their first question, start right from the day they join and tilt it. And we call it Good Day because, of course, we're from Australia. So you have to. And what happens, so what we have learned, because we have been a power user for a very long time there, all our internal process flows are being built by HR in that.

[00:19:36] So that has been a very, I would say, we are a little bit ahead, unlike any other HR team, because you have to rely on IT to deploy some of these processes. So today, if I need to deploy our promotion process and promotion nomination process and make changes, I can do it myself going into Jira service management. So we do that. Our AI-enabled virtual service agent that is sitting in JSM, it handles, I would say, 25% of our query that comes through it. And we have learned a lot in that journey.

[00:20:05] It wasn't 25%, I would say, three years ago or two years ago. Currently, we get about close to about 4,000, like 3,800 queries per month in this thing. And we have learned a lot in terms of how can you reduce those queries coming in upstream. So don't track like, hey, how much volume you're solving. It's not a good thing. The lesser the volume it is, it's actually good. And how much of that volume is being solved by your AI-enabled agent is what you should be tracking.

[00:20:32] But I think in all of this, one thing that we track is our learning has been to not make all of these adjustments in order to get efficiency. I think that is my biggest learning. And that is my one piece of advice to everyone who are power users or going on the journey. Don't start from the economic value to say, I'm going to save X number of hours or I'm going to get this much of savings from A. I know every company needs that for a business case. You need that. Fine. You can do it for that purpose.

[00:20:58] But when you're actually implementing it, don't have that as the end goal. Have the customer satisfaction score as the end goal. So we have maintained 95% as our customer satisfaction score to make sure like, hey, we cannot drop that. Whether you're using it in HRSM or you're using it in CSM. So once you do that, then it's a very different implementation that you will go through. So we have maintained that. So my one learning I would say is like, don't solve for the economic value.

[00:21:27] Solve for the customer satisfaction or whatever is that index, which is about, that should be your desired outcome. And second thing, I don't have a good answer for it right now is that when people are using AI, and I think this is going back to my earlier point, when they become strategic collaborator, they do get extra capacity in their time to do some deep thinking work. Everyone is trying to quantify that, to say like, fine, Apni, now this person is using AI and it's a super user. And they've got 20% extra time to do other things.

[00:21:57] What is that other shit they're doing? Like, how do I quantify? Don't rush to do that. Because having been on the journey myself, I'm speaking, when you get that time, you get more creative. You're thinking long term, you're thinking on the next strategic initiative, which is yet to come out. So let that, don't rush to just quickly monetize what that saving is of the 20% that I've saved.

[00:22:24] It would be my other advice right now. It will show automatically. Let your one employee come up with the next innovation and idea. And then you'll be saying, oh, wow, like this happened because I freed up this employee from doing this work. And I gave them 20% capacity. And speaking on a similar theme there, one of the things I try and do at the end of every podcast is give everyone listening a valuable takeaway. And I think the one universal problem that we'd all love to solve is trying to find information.

[00:22:51] I think you and I have experienced it. Everyone listening, we're working in silos. And somebody will come and ask you for a PowerPoint presentation template, a tech note, a SharePoint document, an email sent in February last year. And we can't find that information. So with your focus on building world-class distributed teams, I'm sure you've come across this many, many times. So can you give me a peek behind the curtain into the kind of tools or practices that have been most useful to you and your team so far?

[00:23:20] So first of all, we don't use emails. Like, so I haven't. Yeah, so I'm not on my emails. So you have to really change your tech stack dramatically to do that. So that's one. Second point is meetings, like I said, don't replace meetings. We have replaced close to about half a million meetings in our last two years by using Loom alone. We recorded about 14,000 Loom every month. And it's not just about the number of Looms. It's the activity and the engagement.

[00:23:50] And my one experience of using Loom is I'm an introvert myself. You will see the participation in a Loom, this thing is way more, it's richer. It's a richer conversation than a meeting because people who would usually not speak get to stop, pause at 2.20 seconds and say, Hey, I don't think so. This is what is making sense. I would say this is how I would do it, right? You never get that opportunity in a real life meeting. It's even worse when it's a physical meeting.

[00:24:18] Everyone is reading the room and the body language and all. So it kind of just removes all of that biases. So I would say the quality of meeting, it saves the time. It's much richer. So that's Loom would be another thing. Third is how you structure your day. So make sure that you tell your team like, Hey, you only need four hours in a day to do collaboration, which is meeting time or whatever you're doing. Remaining four hours of your day has to be deep work time where you're actually doing some deep thinking and then you're executing on it. So how do you structure your team?

[00:24:48] That is how you structure your day is like one very strong ways of working this thing and try and move to asynchronous ways of working. And it's a different way of thinking, but that would be my advice. And of course, with our over-reliance on technology and doing less meetings and maybe less face-to-face time, how do you ensure that you don't lose that human connection in the workplace? How do you get around that? That is such a great point. And I'm glad you asked.

[00:25:18] A, we are distributed first in our thinking and we are saying these are the ways of working. At the same time, we also feel like what makes it happen is the human connection. I also want to be, I want to clear the myth that connection is built by coming into office or connection is built by sitting next to each other on computer and looking at your own screen. There is no connection being built. I've heard from my customers and my peers that people do town hall in an office setting and people are sitting on level one, level two, wherever, wherever.

[00:25:46] And some of them are dialing into the town hall, I assume. So it's a myth that connection is being by having people in the office. I would say connection is based on our research and our experience again. Like I just want to make sure that it's based on our own learning. Connection is built when you get people together to solve a very difficult problem, right? Or give them a strategic initiative to go and work on. So it has to be intentional to start with, which means you create intentional moment of togetherness.

[00:26:13] So if you are a design team and you have to do a brainstorming session, you're working on your next thing, 100% bring them in together. They want to whiteboard. They want to spar on ideas. It needs some creative thinking. Do an ITG, right? We call them intentional togetherness. An event like this where we are talking over right now, we've got our customers in, we've got our partners in, we've got employees in. Do you create these moments? So this kind of interaction that you have builds much stronger connection.

[00:26:38] Because you're here to solve a problem, not here to come into work because A, you've been asked to come two days a week or three days a week. So it's a very different emotion. So I just want to make sure that that's how we do it. So we get teams together based on whatever works. If you're an engineering team, once a quarter, once whatever the frequency works. And also get each craft autonomy because it may look different for HR versus finance versus engineering. And then bring them in together to really solve a business problem and be intentional about it. Wow, that is a powerful moment to end on.

[00:27:07] But before I let you go, we're recording this on the last day of the event. When you get that plane ride home, what are you going to be thinking about from all the keynotes, everything you've seen, everything you've heard? What are you going to be reflecting on that plane ride home? If you're talking about it, this has been a power week for all of us in terms of got really great insights from our customers. So I'm going to go, I have an action list. And like at last time, we want to move really fast. I'm going to be working on that on my ride on the plane back to Sydney.

[00:27:37] And second thing is I've heard so much excitement from customers about Rover being free. So really working on in terms of how do we help them on the adoption and how do we help them and work alongside to say like, hey, this is how you play with Rover. So those are the two things. Awesome. Well, I will add links to everything we've talked about today so people can find you nice and easy. I love to see how this journey continues to evolve. But just thank you for talking with me. Thank you so much.

[00:28:03] I think as Avani reminded us there, it's not just about rushing to quantify AI's return on investment. It's actually about giving people the time, the headspace, all the things that are required for us all to think bigger, work smarter and contribute more creatively. And the Alassian approach isn't just built on buzzwords or sweeping mandates. It's built on listening, experimentation and empathy.

[00:28:28] Would it be building internal tools or rethinking how and when teams come together? It seems that they're proving that scaling technology doesn't mean sacrificing human connection. It's more about enhancing it. But what about your own organization? Are you embracing AI as a top-down initiative? Are you empowering individuals to explore the possibilities on their own terms? And how are you keeping the human at the heart of your digital transformation?

[00:28:58] As always, love to hear your thoughts. Drop me an email, techblogwriteroutlook.com, x LinkedIn, Instagram, just at Neil C. Hughes. Love for you to share your perspective with the Tech Talks Daily community here. Other than that, it's time for me to go now. I'll be back again tomorrow with another guest. But keep your messages coming in and I'll speak with you all tomorrow. Bye for now. Bye for now.