Cloud Reset – The Podcast | Episode 12: From Azure to Agents: Building an AI-Ready Business with Rachel Bondi and James Mystakidis
Show Resources:
Here are the resources we covered in the episode:
Follow Naran McClung on LinkedIn
Contact us at enquiries@macquariecloudservices.com with any questions, suggestions, or corrections!
And don’t forget to subscribe, rate & review!
Episode Summary:
In this special episode of Cloud Reset, we reunite the original architects of Macquarie’s Azure journey—Rachel Bondi, Corporate VP SMB Asia at Microsoft, and James Mystakidis, Group Executive of Macquarie Cloud Services—for a clear-eyed look at what it takes to build AI-ready organisations and cultures.
Hosted by Jono and Naran, the episode revisits how the partnership got off the ground and explores how those lessons are shaping their approach to AI.
Inside this episode:
- How repeatable IP, obsession with NPS, and product innovation built customer trust at scale
- Why every successful AI project starts with data (and the case for Fabric + Purview in the mid-market)
- Behind the scenes of Macquarie’s new Customer Insights agent platform
- What it takes to drive AI adoption—securely, practically, and culturally
- The rise of “agent managers” and how leadership roles are evolving
- Why Copilot isn’t just a tool—it’s now table stakes for attracting top talent
Rachel and James don’t just talk about what’s possible—they’re building it. If you’re a business or tech leader wrestling with how to start (or scale) your AI journey, this one’s for you.
Watch or listen now👇
#CloudReset #AI #Leadership #DigitalTransformation #TechStrategy #MacquarieCloudServices #Podcast
Episode Transcript:
Okay Jono drum roll. We are back. I am Naran McClung, the head of the Azure business for Macquarie Cloud Services. Who are you?
I’m Jono and I look after private cloud for Macquarie Cloud Services.
Very good. Lovely to have you back, Jono. This is of course another episode of the Cloud Reset podcast, but before we get to any of that, and before we talk about our guests, which are both amazing people in their own right, what on earth have we been up to?
What have we been up to? Well, uh, I reckon for the last three or four months we’ve been desperately trying to avoid the Q4 madness. Unsuccessfully.
Is this the hockey stick? Is that what you’re referring to?
It’s upon us for some reason. Everybody wants to get everything done in Q4 of the Australian Financial year, right?
I’m not complaining. I’m not complaining.
Now, hang on a minute, Jono, this doesn’t seem to work because the horrible things are going on in the us. It’s affecting the stock market here in Australia. Surely that must have had an impact on market confidence.
Well, uh, look, depending on who you talk to, it’s possible, but certainly not in our world.
I’ve seen no sign of it. No, uh, no sign of slowing down. Things are getting crazy. Crazy.
Absolutely. So I look, you know, we both know we’re having fun with this clearly, but businesses are on the move. We’ve got huge enterprises that are moving outta data centers. We’ve got equally big enterprises moving out of AWS.
I talk about that a lot. For my own ill-gotten purposes, but it happens to be true. All of these things are happening at once. Why do they all. Seem to land in Q4 of the financial year.
Well, I think, you know, there’s a little bit of pre-Easter madness and people wanna get some stuff done before they go on holidays.
Right. I think Australians love a good holiday.
Yep.
Um, and uh, and then as we head into May and June, it just, this has been true, uh, you know, for at least the last two or three years. Mm-hmm. People putting things on hold and then, and then a bit of a panic.
It’s never gonna be any different, is it?
No, I don’t think so.
That it’s never gonna be any
different at all. We always like things to be smooth and linear throughout the financial year. It’s just not the case.
No, but look long mate. Continue. Let’s get crazy. I wouldn’t have it any other way.
No, indeed. Let’s go.
Let’s go.
Okay, here we are. We’ve got two special guests, both of which very close to me and my own origin story in working for Macquarie Cloud Services. First of which we have a lady who originally was the Chief Partner Officer for Microsoft when I first met her. She’s now the corporate VP for SMB Asia at Microsoft.
Her name is Rachel Bondi. Equally, we have, uh, our boss, Jono, uh, the group executive for cloud services. James Mystakidis. Welcome, Rachel. Welcome James to the podcast.
It’s great. It’s a, it’s a high bar, having your boss on the podcast. No pressure at all. Talk us through, how did you guys meet? Ah, I remember it like it was yesterday, but I think it was six years ago.
We had a business meeting because, um, Macquarie wanted to expand into cloud technology, and I was just new in role and I was like, wow, who’s this partner who wants to meet with me and talk about growth of our business? How can you say no to that? So I remember the first meeting we were in the city. It was such a well thought through business plan.
It was really clear on what you wanted the partnership to be. And at the heart of all of it was all about mid-market and customers. Um, so it was one of those moments to kind of look around and say, this is just too good to be true. And that was six years ago. Wow. And the partnership has just gone from strength to strength because one thing I’ve really appreciate about Macquarie is when you say you do something, you do it and you deliver it with such excellence.
And every year, year after year, you’ve raised the bar on what we can do together. I’m really keen for you to expand into Asia, and I know it’s been something we’ve been talking about, but I also know that’s what makes you so great is that you stick to your core. And maybe the last thing I’ll say is, um, at first of all, this started around the practice of Azure.
Yes. But it became so much more than that. It’s actually where it goes from here is probably the most exciting part. So I know you’ve done a lot more in security. I know you’re now looking at data and fabric and who knows where this will go when we think about AI in the future. But No, I, I remember that day so well, six years ago, and I’m so, so thankful for the work that we’ve done together, and I think the, the next six years are really gonna probably be the, the biggest of, of all of them.
I see. Thank you.
You’re right. We, we have a, like a track record, a process around how we innovate. We normally, uh, travel abroad, um, and we go to regions where we don’t compete and we talk to companies just like us. And we do that normally on an annual basis. And I remember 2017, 2018, we had done two study trips to the UK US with the question of, you know, this public cloud thing, is it for us?
Yeah. What are the tricks? What are the traps? Who do you partner with? Uh, what’s the right strategy? How do you make money? How do you offer great service? And I, think we came from that convinced that we needed to work with Microsoft. Yeah. Really convinced that. In fact, they’re the only partner we could work with.
Um, lots of reasons for that. If we think of the segment that we target, mid-market, they’re mostly Microsoft customers and our view was Microsoft is the only hyperscaler that’s partner first. Yeah. And they have, uh, decades of working with other businesses to service their base. So I think they were the two reasons.
We knew that we needed to then start to open dialogue. And I think we were lucky that we met you. ’cause I’m sure you know, there’s been lots of partners that have come before promising big things, uh, and all this hypergrowth and you know, it, it doesn’t turn up. So our, our view was like. We’re actually really competent.
We’re really good at what we do, but how do we not be the thousand and first partner that turns up promising things? And I, I think I recognised that, uh, you had the insight, you asked the right questions. I think you triggered on our obsession for customers and net promoter score journey. And I was like, okay, here’s someone who actually gets
how we’re different. Yeah. Why we would make a difference and why we’re gonna have success together. So. That was, that was my story.
It’s a great story. And look, can I tell you, um, being hired to, to lead the Azure practice some six years ago and inheriting, we talked about study trips. I think that understates the effort that you put into it.
It’s such a privilege to inherit all that work. I mean, it was two years of study trips to America and to the UK, I think 40 something odd vendors that you met with, and it was an astonishing amount of information that I received. Um, I’ve never had anything like that. I’ve probably never had anything like it again as well.
It’s been incredible.
Yeah. Look, I think it’s a good journey. Um, that’s really how we innovate. You know that now, now Naran at Macquarie, we do it on an annual basis. Mm-hmm. We go, markets that we don’t compete with, we talk to people who are just like us. Mm-hmm. We share them the insights about what we’re good at.
Mm-hmm. Share them our, net promoter score journey. Our, uh, focus on people and culture. Mm-hmm. And then in exchange, they share with us their insights about what they’re doing. Sometimes the markets are ahead. Of us sometimes they’re where we’re at, it’s a great opportunity to look into the future sometimes bring those learnings back.
Yeah. Uh, into this market.
I think when people are doing a good job, they’re proud of it.
They like to talk about it and share. Um, so as long as you’re non-competitive, I think. We’re all proud of what we do. Right? Yeah. It’s a, it’s a great way to learn.
Rachel, we’ve been, um. Obviously doing a, a series recently on this podcast around embracing new technology trends. Um, the one, the one we’re all talking about is AI, but if we, if we go back to when I think Macquarie, uh, commenced the, you know, the Microsoft Azure partnership, we’re going about six years now.
Um, that was kind of a new trend as well. You know, you’ve been going for a couple of years, but, you know, hyperscale adoption. Uh, certainly in Australia was, uh, was at the, at the beginning, you know, and, and there was a real wave and it feels a little bit, you can draw some parallels to adopting new technology trends.
What stood out to you when you engage with Macquarie around. Macquarie’s view of, you know, how to, how to get on board with a new trend. What, what, what stood out to you as I important for that to be successful?
I think the, the thing that I noticed, um, with Macquarie, um, right from the onset when we were talking about Azure, is this concept that you could bring your expertise and experience to other customers.
And that for the market that you serve and you serve so many different industries and you serve many different customers, a vari variety of different sizes, you wanted to instill confidence with them on what we call sort of repeatable. IP repeatable, understanding the best potential experience, and you have packaged it in a way that makes that migration to the cloud even more seamless for customers.
Yep. So I think the combination, again, of you being really focused on net promoter scores, so customers ultimately saying this is how you’ve supported them on the journey, as well as bringing the experience that you have from many different customers to the practice is what I think really, really stood out for you.
And I must admit, you know what you did in year one, is completely different to what you do now. True. The scale that you have now. And I think that was another part of the journey, was that you wanted to make sure that you had a year one plan, a year two plan, and a year three plan. And you would only take on what made sense for you at that point in time.
And I think for me, it goes back to that point. Customer feedback is absolutely paramount to how you do things. Yep. And then making sure that you’re taking the best of your learnings to other parts of the business. But I dunno if James, if that’s was intentional or how it ended up being, but that was what struck me as how you took customers on their journey to cloud.
Well yeah, James, I’d love to hear more from you on that around adopting new technology trends, ’cause that’s what our listeners are really interested in. But there’s also another, uh, a follow on question that I’ll prep you for, and that is, uh, as a leader at Macquarie. A foreign new technology being Microsoft Azure, you know, the change management around that, and how did you bring the rest of the business on the journey?
So obviously you had a great plan, but execution is also really important as well. And um, I’d love to hear more, more on that. Yeah. Got it. John,
I, I think, um, something I’m deeply passionate about, I think your question for me touches on our purpose. Our, fundamentally, we believe that the market, the mid-market business is overcharged and underserved.
Um, and that’s not true for the enterprise or consumers. And in the enterprise, they’re not overcharged and underserved. They have massive teams and procurement offices, and they can negotiate in the consumer space. Every service out there has 30 alternatives from insurance from, internet, whatever you want.
That’s hotly contested. Mid-market are subscale. They don’t have the people, they don’t have the teams, and therefore they find it difficult to service themself. So to, to Rachel’s point, our business model is structured around how do we serve mid-market on a long-term and ongoing basis. We’re interested in short, quick projects.
We’re interested in long-term, uh, relationships. Mm-hmm. Uh, that continue to serve them. And we, we, the burden is on us to drive value and ongoing value through the lifetime of the relationship with our customers. That’s the burden we have.
Yep.
Uh, and we work every day to win, win the right, the right. To serve those customers.
I’d have to say that’s one of the things that really struck me about Macquarie. I mean, you’ve raised the NPS piece, that’s super important, but the feedback goes into product development. I’ve never seen anything like it at Macquarie. I mean, our, our product managers produce 60, 70 pages of service description that manifests in automation and capability and product as we serve Azure customers and Microsoft security customers.
And that commitment to product goes to not just attracting customers, to, to your point James, it’s about retaining them as well. And we have a real commitment to hanging onto a customer. Hard fought customer one needs to be held onto. And I know every month, every week, we’re absolutely committed to making sure that our customers feel like they’re getting value.
And it’s funny with NPS, if you get anything other than a promoter, it’s like a red light goes off in our business, right? And everybody panics and like, what’s going on here? Right. And, and, and it’s a good thing, right? And our customers know it too. They also know that they get attention as well. And typically it will be something that’s important to talk about.
So it’s a massive part of our business. And that whole commitment to product still strikes me. It really does. I like, I don’t, again, don’t think I’ll see anything like that. I, other than what I’ve seen. No.
Look, I just to pick up on two points that James has said is certainly the focus on mid-market. There are so many partners.
Who tend to focus on like the, the largest of the large, and that’s what makes Macquarie stand out is this whole market where I talk to so many CIOs and CEOs, and I’m sure we’ll talk about more advanced technologies coming. Shortly, but they want more support and that’s what Macquarie can provide them.
And then you’re absolutely right on the investment on product management or product marketing. Um, I don’t, you know, I don’t see many other companies in Australia doing it quite that way. And it’s refreshing ’cause that’s how Microsoft works. We tend to work very closely with our engineering teams. We work and invest in product marketers ourselves.
To make sure that we are really articulating the business value that you get from the technology platform. And that’s what I think makes this relationship work so well is that you think like a technology company and that just continues to deliver great value to the mid-market of Australia. Yeah, it’s fantastic.
Yeah, look, it’s amazing. It’s, um, it’s a privilege. It really is to be. Part of this business. But anyway, I don’t wanna get into that. I’ll start welling up talking about that. But we are talking about waves here. And Jono, you raised the 6 years ago, Azure was a different beast. And my God, there are so many different services now.
The platform has evolved immeasurably. The focus on security, uh, and everything else has been incredible. It was a wave of itself. Watching Azure evolve over that time, we are seeing a new wave now as it relates to machine learning. And generative AI and Jono. I know we’ve been speaking to our guests a lot about this in recent weeks, and there has been some.
Consistency of insights coming through with that?
I think so, yeah. A lot of, a lot of consistency in terms of, uh, you know, the people that we’ve been speaking to around how to adopt AI successfully and make a difference for their own business or for the customers or markets that they serve. Um, and something that’s sort of come through in, in both of your feedback around how do you adopt a new technology trend?
How do you make it real in your business for your customers? And I’m hearing a lot about linking it back to your purpose. Yeah. Right. And when we’re speaking to other tech leaders talking about. Adoption of AI and you know, which things do we adopt and, and where do we focus and how do we get return on investment?
I think they could do well to think back to, well, what is the purpose of our business and the market that we serve, and how do we make a difference? One of the common themes that came out was having opinions based on lived experience with the technology. Right? Actually getting your hands dirty with technology as business leaders when it comes to AI and having an informed opinion about what
good looks like for your business. Um, do you have a point of view on this, uh, Rachel around, you know, Jono, it’s such a, such a big topic. It such a big topic. It’s huge and well, and, you know, Microsoft’s leading the way on this.
Yeah. Look, it’s, it’s a great topic. Um, I, I remember when I first started in the industry and had the opportunity to, and I’m dating myself here a little bit, Jono, but when I was first using sort of word in Excel, like I was truly.
Like it really amazed me. I have never quite felt the same as I did that time. The only time that has come close is when I started start using ChatGPT as part of Microsoft Services, the ability to have natural language responses. Using prompts to help in any type of summarisation. Creation has been a game changer.
So from the moment I started using the tool set, you know, two and a bit years ago, um, available through sort of the Microsoft technology set, it has been an absolute game changer. I think for me, you cannot stop learning to begin with and it has to start with leadership. You know, I spend a lot of time myself talking to the team.
Trying to spend time with myself, just understanding and learning sort of the technologies and I, I use it every single day and as I sort of go back over years, it’s completely changing the way in which I work and support the team as well. So what are we doing? Sort of, you know, our organisation, you know, make sure that the technology is available to everybody in the organisation.
I think that’s job number one is to make it broadly available now. I would certainly recommend in our organisation, we give it available and Copilot available to everybody in the team, and we’re also allowing the teams to create their own agents now. So through the use of our technology called Copilot Studio.
People can create and build their own agents, and that has been something that I think is the next evolution and where we need to go to. So yes, people are using it as an assistant, which is what I started using it as. But now I’m using it to automate so many different business processes so that myself and the team are focused on
the things that we really wanna focus on and, and eliminate some of the business processes that we think are not as high value. So making the technology available to everybody in the organisation is so, so important. So that’s sort of step number one. And I, I do wanna get James’s thoughts on this. I’ll, I’ll, I’ll maybe tee him up with something.
Before you can give people access to the tooling and the technology, you have to do it in a way that is secure and that the data is available. Um, now the security piece is so, so important because you wanna make sure that people can’t search on data sets that they shouldn’t be able to get access to.
And then the data piece, that’s an ongoing component because initially in some of the search results or some of the, the prompting results, it may not be exactly what you think it should be. And that’s where you have to keep on giving feedback to the tech teams to make sure that they’re incorporating the right data sets.
Um, so I think, again, giving access to everybody in the organisation to the tooling, and then also ensuring that you have the right security and data pro protocols in place. It’s so, so important as a business leader. But no, James is doing a lot of work too in this area. So I’ll pass to James for his thoughts on this topic.
Well firstly I’d say I’ve seen AI, I think fit broadly in two camps. Um, firstly, I’d say by no means do I see myself as an expert in this space. I’m learning along the way. I think the first way I would call it individual productivity. So there’s been lots of tools like Copart Chat GPT, that help individuals do things cheaper, better, faster, more efficiently, save time.
And I think that’s really good. And I’ve certainly been leaning into those tools. My observation is I’m surprised how much they’ve evolved and how, uh, complex they have become and how. Much value they’ve, they deliver now compared to one or two years ago. The second thing I think is around organisational efficiency.
And that’s where I’m personally very interested in how do we create significant efficiencies in our business that help us scale, uh, frankly to provide a better service to our customers. Um, I’d say we have three projects within our business that have, uh, mature. We have one around our security operations center.
We’ve built a whole Microsoft security practice on Sentinel with Microsoft. We have significant AI, machine learning automation blueprinting. We take in threat feeds. It’s real time analysis. And what does that mean for our customers? It’s machines are now picking up and identifying threats and also activating responses without having to have our people involved.
That then triggers our individuals to then look at what’s happened. That means we can get to near real time response, which was unheard of only a few years ago. Uh, the second areas around our hosting management center. We use it for ticket enrichment for service management. Uh, but the project that I’ve been personally leading for the last year and a half is around customer insights.
And this is how do we help our organisation, our staff, better improve our engagement with our customers and create material efficiencies. And if you think of an account manager, they have say 50 customers and at every customer they have five contacts. How do they keep track of 250 people every day and their next action and their contract resign and have they changed company and when did I meet them last?
That’s a significant opportunity to drive efficiency, automation and help the tools and the systems, uh, support and be the agent for our service managers. And what we discovered when we embarked upon that project, we had this idea some time ago. We couldn’t do it. We, we quickly figured out that we needed to run a data project to run an AI project, and my sense is, is that most organisations, particularly mid-market, I think Enterprise are far more mature around their data and data sets.
Mid-market uh, most of them aren’t data-driven businesses and they don’t have a data lake. They don’t think about data governance. They don’t think about security. They don’t think about how much data to store, uh, who’s accessed the data. Uh, I think there’s a lot of, uh, headache, uh, and learning ahead. And so we had to tackle that first, frankly.
Now, one of the things that, uh, tickled my brain was how can I use AI? To not to look at data in a different way. Can I skip a data transformation project? So that’s one of the technical challenges that we took on, and I think we’ve made a lot of progress there. You know, by getting a, a copy of every one of our data sets into a one lake and then now it’s there.
How do we write custom agents that can understand each one of those data sets to curate? Make better sense of that data that we can then serve up to, uh, other agents to answer questions. So that’s a bit of the journey that we’ve been going on and I think some of the challenges ahead talks to Rachel’s point, uh, which is I think data platform, data management, data security is gonna be a significant wave ahead.
Many organisations are gonna need help and support in that. And that’s really the question where. Tackling right now. How do we build a managed service around that to support our mid-market customers on that journey? Yep. Yeah.
Yeah. I was just gonna say like, uh, when things have a natural gravity, I love that in the industry and fabric for us.
I mean, we’ve spoken about this a few times now. Um. Has that gravity and we are being drawn to it and our customers are embracing it and they feel compelled James, to address their, their data platform problems upfront. They recognise that that’s crucially important. I think data governance is a fascinating one.
Rachel, you know, you talked about, I. The risks potentially of models exposing data that shouldn’t. Right. And so changes in that space represent cultural changes as well. And you change the way people work. We’re all data domain custodians. We create documents. We have to think about how they’re classified.
Anytime you have to interrupt someone’s flow in the way they work, it takes time. That’s a real cultural change. And I think businesses are going through that right now. But fabric itself has really exploded in our business. And I would say not just customer demand, it’s peaked the interests of our.
Engineers and our mad scientists. And you are one of those, James, with your customer insights project. And I say that affectionately you had an aspiration to solve a problem. The technology was super interesting. Um, we’re evolving our own data platform story, but I love that you had a, um. An alternate strategy, um, to address that with agents and let’s not understate it.
There’s like 20 different models in play with that solution and so it’s pretty elaborate. It’s pretty cool. Um, I’m looking forward to working with it as well. I know the account managers in our business are as well, so it’s very transformational for us and it’s gonna change behaviours in our business.
It’s a really, it’s a really good point because we, we have a similar, um, agent that is available to our sort of account managers.
And I think for me, and, and James mentioned it, is people want to be able to support customers more. You know, there’s not a, there’s not a great, a great account executive out there that doesn’t want to serve and provide additional value to their customers. But how do you do it at scale and with accuracy is, is where I think the big unlock is.
So to me, it all comes down to. You know, using AI to unlock human ambition and that it’s not to replace one or the other, but it’s to really assist and to provide, you know, more scale in how we, we serve our market. So between the work that you are doing and the, and I know we have a similar type of setup and we’ve just rolled it out today, in fact, to all of our sales teams, um, across the business in, in our, in how we’re working is how do we continue to have this human and assistant, um, through an agent.
Sort of approach. And I can imagine that our account teams in the past will have multiple agents that they’re having to manage. And that’s a really, really exciting, uh, venture, but also one that’s gonna require a lot of change management from the leaders all the way through into the organisation.
It’s one of the things that Michael Reid is the, um, the CEO for Megaport suggested that perhaps a change in his leadership team was that there’d be leaders whose job it is to manage agents.
Yeah. You hire an individual who has agents underneath them that deliver outcomes and that is their job.
Absolutely. Like I would imagine that, you know, your account executive in the future will have four or five different agents that they’re having to manage.
📍 Which is a good segue. So there’s two stats. So firstly, apparently 47% of businesses in Australia don’t feel like their data platform strategy, um, has been realised.
So that doesn’t really surprise me that much given the interest that we’re seeing with fabric. That’s the first thing. Second thing is adapt research suggests that investments in gen AI projects have now overtaken investments in security initiatives within organisations as well, which is interesting. I know the two go hand in hand indeed.
So it’s not a case of one or the other. So I guess my question would be. What advice do each of you have for businesses that are seeking to champion and adopt AI adoption within their own business? And how do they be mindful of those other investments? How do you tackle that?
Okay, I’ll, I’ll make a comment.
I think I’m, firstly, I’m not surprised by the first stat. Um, I think that the AI technology is ahead of where data platforms are at, right? I think that is the next biggest challenge ahead for businesses to get their data in place, in shape with the right security, with the right governance, with the right accessibility to create insights.
I think the agents that we’re working with today are very advanced, but have not as much data to work with. Okay? So I think that’s gonna be a very significant challenge. Now here’s the second problem with that. No one wins an award. Or Ends up in the CIO magazine for getting their data platforms right.
Right. Uh, and boards and, and businesses aren’t willing to write and invest big checks to get data platforms. Right. Okay. And I think that’s the really a massive issue ahead. And I think, uh, CIOs and businesses, uh, that traditionally haven’t been data. Focused are gonna need to think about actually how they figure out what their investment in data is first, and then think about what AI outcomes, uh, that they’re looking for.
Yeah. So I think that’s a bit of an advice. There for the CIOs, uh, had now our view is, at least for mid-market, most of, uh, these customers are Microsoft customers. Mm. So. Get your data into fabric. Yes. Get your data into, uh, one lake. Establish purview. Establish the data governance, establish the security, establish that is your very best place to start your data journey.
At the very least, you’ll end up with the bronze. Data medallion architecture. Yep. And from there you can start to think about how you can get to silver and gold and what your drivers are. Yep. I think that’s, that’s the advice that we’re, um, providing to our customers and we wanna help them on that journey.
Makes perfect sense. In fact, talking to our, um, security analysts as well, I said to them, um. What would be the number one weakness that you see across all the customers that we manage? And the universal answer to that from all of them was data leakage prevention. They’re just seeing real, immaturities there, which goes to this, this marrying up of your data strategy and your security strategy.
And so their suggestion would be, as you say, James, embrace products like purview that can help with that. Um, but also importantly, help with the data governance, sorry as well.
And look, I, I. I completely agree with, um, what James has mentioned. I think the other thing I would, I would ask folks to do, because Australia, many, many Australian companies have already invested in Microsoft 365.
So once and, and also there’s probably a stat out there that says how many people are already using AI tools that are not endorsed by the company, right. So my recommendation to anybody that is using Microsoft 365 and in Australia, it’s the majority of our customers, is to enable Copilot chat. It’s gonna use all of the data that is in your office graph.
It’s going to held within the tenant. Has the right secure protocols and it’s in, it’s available for most customers that have already made that commitment to Microsoft 365. So I think for me, different AI capabilities will unlock different things. What James talked about is certainly going to help unlock business processes, customer intelligence, but if you wanna just make a, a
first step into employee productivity and employee experience, leverage copilot chat as part of the Microsoft 365 and, and make sure that then, you know, employees aren’t using third party tools that are not endorsed by your company.
Yeah.
Um, that’s how, that’s, that’s the type of thing that a CIO doesn’t wanna get into trouble for, which is, you know, you are using, using tools that could allow the company data to appear in the public domain.
You had an interesting take on this when we were chatting the other day. Slightly contentious. ’cause I know Naran and I challenge you a bit. We say okay, I don’t know, we talked to a lot of mid-market companies and they go co-pilot’s, a bit expensive. Yeah, we’re not sure. It feels like a bit in that camp that James spoke about the, the assistant type category.
Um, and you were talking about chat GPT Premium subscription shadow it, I think, and the comment that I took away was, well, what’s an hour? Back of your time worth to you?
Oh, look at scale. You know, I, I think and
pricing models for this stuff.
Exactly. Like, I think overall, um, I think this is a place where there is going to be, you know, business model transformation.
You know, whether it be through how, you know, you use search on, you know, on browsers all the way through to how you think about productivity in the future. But I, for me, it comes down to. How much can this unlock in terms of productivity and employee experience? We’ve had some people say to us, um, when they’re attracting interns or, you know, new recruits into the business, is that some, some people coming into organisations are saying, do you enable AI as part of your company?
It used to be, do you have a pc? This was back when maybe James and I started working in the industry. Now it’s like, I love that she excluded us from that, by the way. We’re younger. We’re younger somehow.
Yeah,
that’s, thank
you for that. That’s okay.
But now people are saying, you know, do I get AI tools if I join, if I join your company?
So for me, I think it’s something that a lot of employees, it’s an expectation and what the value. It pertains, it’s, you know, a dollar a day savings, $5 a day. Like, it, it, you don’t know exactly at this point in time what is too much or too low.
And pricing models will continue to evolve. Absolutely. What, what might it look like in the future?
Look, my personal opinion on this is, you know, subscription models today have always been more one of per user per month over that sort of timeframe. But maybe we move to a model that it’s more outcome driven. Yeah. You know, you’re paying for, you know, somebody to help you, um, you know, manage, manage your calendar or Yeah.
Um, somebody to, to, to book you, you know, a holiday. It might actually be that you’re paying for a flat service. As part of that offering versus the per user, per model. I like that, um, approaches that we’ve had. So I think the way in which this will evolve into driving outcomes, I think that’s the big thing is that this, this technology will Dr.
Drive more personal and business outcomes. And I think how much and how people. Will we’ll monetize that will also change in the future. And
it’s like, um, it, it was a factor for us in the way we architect our own solutions, um, in that we know that some, some solutions can be token heavy, some can be token light.
And we’ve been fascinated with just the ability to eke out an amazing outcome. And if I look at our soc optimiser, our digital twin, we’re servicing just on average, just in this one instance, let’s say a hundred customers, it costs us less than 500 bucks a month for all of them. To have a meantime to respond of like seven minutes and a meantime to closure of under three minutes.
That’s an astonishing outcome for almost no money because we get in, we get an outcome, and we get the hell outta there super quick. And so I think there’s, there’s some artistry in that. Certainly Jono, for a managed service provider, how do you induce a, a really solid TCO for a generative AI project that’s really cost efficient?
We
we’re hearing this a lot now, this idea of Agentic Workloads, agentic outcomes. People talking about AI agents. I feel like AI agents are bringing us closer than ever to tech for outcomes in business. I know, James, you’ve got a point of view on this and you’ve been experimenting with lots. We’ve developed lots of different agents and quite a unique model.
How’s it showing up in our business?
Jono in our customer insights platform. I think Naren mentioned this earlier. We’re using about 20 different agents. Um, and let, let’s give you an example. We have hundreds of data sources, uh, across our business that we ingest into our data lake. One of those data sources is net promoter score.
So every time we, um, complete a customer transaction, we send them a survey for them to give us their feedback and give us a score out of zero to 10. Now. So we have a database, an NPS database with all of our relationship and transactional metrics. Now, there isn’t an agent out there that understands net promoter score.
I. So when we say how is it calculated, or when we ask questions about our NPS agents, don’t know what that is. So we’ve had to pick an agent and then tune the agent, and, uh. Leverage context windows. So every time we ask the agent and NPS question, we’re training the agent on what NPS is. So when we said earlier, we’ve got 20 different agents, that’s the process that we’ve gone through for each of these data sources to train the agents on our data.
So we get really high quality responses. Now, on top of that, we have a scheduler agent. What that does is it unpacks the intent of the question and then it farms that question off to the specialist agents to retrieve a result based on what they know. It’s sort of the architecture that we’ve built, um, that platform we’ve builts called Custom Insights.
So we can now add hundreds and hundreds of data sources, hundreds and hundreds of agents, and we get high quality responses back based on the data source. The agent. And the scheduler that sits on top of that. So that’s kind of an insight into what we’ve done now. Now there’s
a nice tangent. I’m gonna ask you a question, Rachel.
I asked the same question of Steven Worrall when was on the podcast a few weeks ago. It’s a question of general intelligence. Yes. And when will we arrive there? Should we arrive there? Feels like you asked different people. Is it today, is it two years from now? And how will that disrupt businesses? Have you got a view on general intelligence?
What did Steven say? He wouldn’t give me an answer. Now, I I, what I threw to him. Because clearly there’s a contract, isn’t there? Right. Between OpenAI and Microsoft that says that should general intelligence be achieved. And there’s a definition of that in that contract that no one’s allowed to see. And I asked Steven for, and he clearly didn’t give me the answer, possibly, does it?
No. I don’t know. But I didn’t get what I, I
I’m gonna be the same as Steven here, I think.
Right. But like how interesting that we’re gonna arrive at that point and that’s gonna change everything.
It is, it is. Look, it’s, um, I think we’re learning, right? And I think, uh. It just, it, it just blows me away how far things have come just in the two and a half years.
It’s, it’s hard to know. I, I know that we’re probably. In sort of that middle innings, right? Like I don’t think we’re at the, the starting point anymore. I think we’re in this middle, middle part of, of what we’re, what this technology can bring the, the amount of models that are now available.
Yeah. You know, and I
know, I know you guys have been a big proponent of the AI foundry and Yes, we have a huge, we we, while we have relationship.
With open ai, we have a relationship with all of those, um, large language models through the AI Foundry, including deeps seek. Sure. I think Sure. Two days after Deepseek became available, we made it through so sure that
Microsoft had a, had a post up there straight away.
Yeah, absolutely. So I think, um, I think Microsoft will continue to, to partner with the industry leading experts.
And I think, I think your, your point on, you know, general intelligence is one that we’re all sort of thinking and there’s very smart people out there that probably know the accurate answer on that one. Yeah. But one that, one that I think that we’re all continuing to learn at. About
My hope with this stuff is I can be so bold is that like I, I love that we have our projects, our projects makes us more scalable and they make us more efficient and they improve customer service.
And that has been our primary focus with security and customer insights and affording our account managers, the intelligence. And I look forward to being able to drive into work and. Like I’m calling a colleague, I’m gonna call teams. I’m gonna get real time information on the way to a meeting, and I’m just, I’m so looking forward to that.
And equally, hopefully one day I can see us industrialising these things and maybe James, that can be a differentiated position and winning customers. That’s certainly a personal interest of mine. Maybe will arrive there as well. I dunno.
I think the work that you’ve done is amazing. Just hearing this story about how you’ve.
Created the customer intelligence and how you have agents now supporting variety of different business processes and how your team is thinking about how they’re gonna change their, their way of working. ’cause at the end of the day, we’re doing this. To help unlock humans in our business. Yeah. And as long as it’s accruing to something bigger than what we do today, I think that’s exactly the place we need to be in.
Look, it’s created a buzz in our business. There’s no doubt about it. The people that have worked with James over the last 12 months have enjoyed doing it. It’s been like a pet project. They, they work out of ours in their own time. They’re super passionate about it. I’m getting pinged on teams all the time.
Look now, and I’ve built this. I’ve built that. Yeah. They love it. They’re proud of what they’re doing. It’s exciting. They never know what’s gonna happen either. Like some of these models, you know, you start playing around with them. To your point, Jono, on lived experience. It’s only through using them that you get a sense of, right, this model lends itself to this agent, to this data source, et cetera.
It’s hugely exciting, right? There’s no way that we could have predicted. I. Where that project was gonna land at the beginning in, in terms of how it ended up, it was an evolutionary process that that changed on a dime. You know, one day there’s a new model and next thing you’re changing your approach and you had to be fluid and agile and dynamic, and that’s part of the excitement.
Yeah, I, I think that’s absolutely right. And, you know, in terms of adopting new technology, I think we, we started the conversation with going back to. Uh, I’d say Macquarie as a business adopting what was a new approach to serving their customers through Hyperscale Cloud in partnership with Microsoft.
Fast forward to today, now we’re thinking about a new wave of new technology hitting the industry, um, based around AI and agentic workloads as leaders of organisations through these kinds of changes. You, you’ve, you know, you’ve done it before. We’re about to do it again. If you could give our listeners or other business leaders out there a, a piece of advice to get this right, you know, how might they best go about.
Bringing their organisation on the journey of adoption of this new tech, uh, what would it be? And maybe I’ll, I’ll start with you, Rachel.
I, well, first of all, you’ve gotta go work with Macquarie. So, and I, I, I say that, that, check the checks in the mail. That’s okay. No, no, no. I say that because it’s actually more about learning from others.
Okay. Why wouldn’t you, why would you work with a partner who’s not? Embarking on this themselves and taking it seriously because everything I think we’ve shared here is we are learning a lot, we’re adapting a lot. Our point of view may change over a period of different time, and you wanna partner who not just works with you on the data, the security, but the change.
Mm. And I think, I think overall, if I had advice to customers out there that you know, that we see every single day in the Australian market is to align yourself with a partner who gets this that’s actually doing the work themselves, because they’re going to appreciate it. So much more, um, and, and share a lot of that knowledge back into your organisation.
So that, that would be, it’s probably best for me to say that than maybe James, but I think it’s really, really important that you partner with people who get this.
Great advice. And James, now you, you’re going second, so you’ve had a bit of time to think. No pressure. Yeah.
That’s a better time. I think that’s right.
I think, um. If I think of mid-market businesses again, and I think of board rooms, um, and CXO meetings, I think that for me the question would be is whenever they’re embarking on a, a project or evaluating a business problem, I’d be asking the question, how could we use AI here? Uh, and I think every current and future project work stream should have an AI work stream.
And I think businesses who aren’t thinking like that will be left behind. Uh, and that would be my advice for businesses to think through. And I think quickly in my pet project at the moment has turned into, I think our understanding of how important data is and solving the data challenges, both in terms of governance, security, risk, centralisation, uh.
Clarity around what’s good and bad data. And I think that, I think that would be my advice. How can AI solve this problem? Uh, where is the AI work stream in this project? Uh, that would be my advice.
Great advice. Absolute gold for everybody listening. Uh, James and Rachel, it’s been an absolute pleasure.
Having you on Cloud Reset. Thank you so much for your time.
I have thoroughly enjoyed it. Thank you both for having us here. It’s been an absolute blast.
Yeah, thank you. And again, we’re gonna have to show that picture, Rachel, the picture you and I took six years ago when we launched the practice. Ping the picture.
There we go. That’s how it all started. Nice production mate. Thank you very much James as well. Thank you guys. Thanks having us. Appreciate it. So look, what a great conversation with James and Rachel. Um, both of which look hugely important to me personally. Uh, Rachel was there right from the beginning, some six years ago.
Clearly, James has been there for longer. Um, we built the Azure practice in that time. Microsoft had been instrumental to that and we’ve seen so much change changes in Azure in market and business adoption, et cetera. In that time, it was a huge wave the last six years. It’s still continuing, thank God. Um, and, but it’s also changing as well.
And we, we see the parallels here with the embracing of generative AI projects and there was many great insights that came through that conversation.
Incredible. And I think our, our two guests have left us with some pretty hard hitting, uh, practical takeaways for our listeners. Yeah. You know, and, and that really stood out to me.
You can’t do an AI project without a data project.
Mm-hmm.
Thinking about your data and perhaps the easy way. Is to look at low-code, no-code options and partnering with people who have done it before. Who, who, or who are using the technology themselves. Yes. Right. So important for, uh, Australian business, especially in the mid-market.
Right. Um, I know James said, you know, CIOs aren’t getting in magazines for doing data projects. Right. We need to rethink. The way we approach what good looks like when it comes to data. Yes. If we wanna be successful with adoption of AI.
That’s right. And look, and why can’t we talk about the fact that it’s cool as well?
It is cool. It is cool, right? I mean the, the mad scientists in our own business and it’s probably wrong of me to reflect on James as being a mad scientist, but look what I mean to say on that. He was keenly interested in the project himself, but so were all the people that worked on it. And Rachel had a comment on that, um, whereby incoming people are suggesting, or they’re, they’re requesting, you know, is, is AI adoption a thing within your business and why can’t it be like that?
Yeah. And thinking about the future where, you know, the modern, uh, worker is gonna be leveraging AI agents.
To help them be, you know, faster, better, leaner, better, cooler, more powerful. That’s right. More legendary.
That’s right. Yes. Uh, I, I can’t wait. I think it’s great. I think it’s here. It’s on our doorstep, so yeah.
Amazing conversation. Grateful to both of them. Clearly. Please subscribe. This has been obviously another Cloud Reset podcast. You can get your podcasts everywhere, so Spotify, YouTube, you name it, it’s all there. Subscribe. We have many more guests coming up. We’re grateful to have you. Thanks for joining in.
Thanks for joining in Cloud Reset. Let’s go. Let’s go.
