Subscribe
Share
Share
Embed
AI is changing how we work, but the real breakthroughs come when organizations rethink their entire foundation.
This is AI-First, where Box Chief Customer Officer Jon Herstein talks with the CIOs and tech leaders building smarter, faster, more adaptive organizations. These aren’t surface-level conversations, and AI-first isn’t just hype. This is where customer success meets IT leadership, and where experience, culture, and value converge.
If you’re leading digital strategy, IT, or transformation efforts, this show will help you take meaningful steps from AI-aware to AI-first.
Graham Link (00:00):
Grass is always green on the other side is there sometimes when you're inside you feel like you're in the fry hang. I think a lot of that comes back to relationships and how we build a culture around people and value people within November. If I think back in my early career, what I learned from that was some of the hardest, toughest projects I ever worked on were some of the most fun.
Jon Herstein (00:19):
This is the AI first podcast hosted by me, John Hurs, team Chief Customer Officer at Box. Join me for real conversations with CIOs and tech leaders about re-imagining work with the power of content and intelligence and putting AI at the core of enterprise transformation. Hello everyone and welcome to the AI first podcast where we explore how organizations are transforming their operations and impact through artificial intelligence. I'm your host, John Herstein, chief customer Officer at Box. Today we're diving into how technology can drive meaningful social change with Gram Link chief Technology Officer at Movember, an organization on a mission to change the face of men's health from its humble beginnings with just 30 mo bros. In 2003, Movember has grown into a global movement, engaging over 6 million supporters across 20 countries, and funding over 1,250 men's health projects in the areas of mental health, suicide prevention, prostate cancer, and testicular cancer.
(01:22):
Graham brings a unique perspective as a tech leader who's proving that AI can amplify human impact at scale. Whether it's supporting the 10 million men living with prostate cancer addressing mental health challenges or helping young men facing testicular cancer, Movember shows us that cutting edge technology and social good aren't mutually exclusive. Today we'll explore how Graham's team uses everything from AI powered code modernization to creative campaigns like their viral mustache generator, all while maintaining the fun human-centric culture that makes November so unique. Let's discover how AI is helping transform and save men's lives around the world. Welcome, Graham.
Graham Link (02:04):
Thanks, John. Pleasure to be here.
Jon Herstein (02:06):
Well, why don't you give people just a little bit of perspective on the organization, but also your role and the scope of your role at Vember.
Graham Link (02:15):
As you mentioned, Vember is a global men's health charity. Originally starting in Australia, it's 23 years ago now, but very quickly grew all around the globe. There's two parts to Vember. There's the movement and the fundraising side to Vember, and that happens every November officially in 20 countries, but it happens all over the world and we have supporters in all corners of the globe and it's a really amazing awareness opportunity around men's health and to talk about our cause areas. The other side to Vember is our health institute and real world delivery of health outcomes that we deliver. So we focus in prostate cancer, dis blood cancer, men's health in general, mental health, particularly young men's health who are very much in the preventative detective phases of mental health and socalled. We do, we have a collected union panel, so we partner with organizations all over the globe using some of those funds that we raised through vember and other funds as well.
(03:07):
My role inside vember is running what we call the tech runs group, so we're headquartered in Australia and we do operate a global platform. So part of my remit is all that fundraising technology, and I think we may talk a bit about that later, but Vember back 20 odd years ago started the peer-to-peer fundraising play, and so there's lots of them in the market these days, but the technology that we started back then, we've continued to build and deliver our epic vember campaign experience that you get when you participate in Movember. But the other side to my role is running the digital technology for enabling the organization and all our employees around the globe as well as those health impacts. So in areas where we are making technology investment in health parts of my team working to that space and support those initiatives and they varied from clinical quality cancer registries through to really fun digital experiences for new adss.
Jon Herstein (04:01):
That's a really broad portfolio of initiatives that the organization is working on and it's easy to understand, Graham, how you would be motivated by the mission of November, but I'm also curious what drives your passion for investing in the technology and AI side of things, which we're going to get into, but aside from the mission, why are you so excited about technology?
Graham Link (04:19):
I started as a technologist Australia University. I did postgrad e-com computing and it's always glued and stuck onto me when we talk about AI and technology. I can say it feels like a bit of a roo reset in my career. It's the first time that I can remember since the dotcom growth and bubble that we've had back in the early two thousands that the rule book's been flicked and at a time where expected to rely on a lot of my experiences and knowhow. Everything's changing around us and so I'm super, super excited about that. John, I think it's an amazing opportunity for all of us in the industry and it's refreshing. There's some real change coming through in the technology here that is allowing us to be better. I think AI allows us to be better
Jon Herstein (04:58):
And we're going to talk about some of those ways, but it's a really interesting point that you're making you expect at this point in your career. I think we've probably been both working for a couple of decades or more that hey, you're an expert, you really understand technology, and then everything changes. We're all trying to figure it out together. Right? Yeah, I just think it's interesting for those of us who've been in the technology world for a couple decades or more that we thought probably at this point in our career we would be experts in technology and really understand the things and be able to pass that on to others. And the reality is with ai, everything's changed so fast in the last couple of years. I'm not sure any of us are experts in anything anymore.
Graham Link (05:30):
It's really interesting, John, if you go back the last decade I've been working with AI with teams around machine learning. The difference that we've seen in the last 18 months, 24 months at tops is just the accessibility of these germs of AI solutions and the technology that's now being built on that and how much it's coming to the workplace. Is it the engineer and business operator level, but all the way up to boardroom just the last six months alone. The narrative has changed significantly in this space for us. We get a lot more questions and seeking to learn at a board level as to how Merera is going to adopt AI and what we are and aren't doing around that and in industry as well. So it's this pace of change along with the potential of the change that's really, really new and I don't think we saw this say as the internet was emerging, we could see the potential, the pace of change was slower compared to where we are today. Yeah, the
Jon Herstein (06:22):
Pace is unlike I think anything any of us have seen before. We'll get into some more details, but can you start with a couple quick examples of how you're using AI already to make a difference on the ground?
Graham Link (06:32):
So the ones that we were very aware of an organizing and then we find out about the ones that have got some grassroots and spoiling that have happened and we sort of have to wrap our arms around quite correctly to understand what's going on in that space, particularly with bigger and more global organization and lot of supporters out there as well. Those that we're really invested in at the low end, the biggest one would be our engineering modernization program. So I mentioned earlier November started this journey over 20 years ago, building fundraising technology. It didn't exist in the marketplace and over that time the platform is being improved every year, every campaign we see new features and a lot of it's being rewritten and rebuilt and it lives in the cloud these days, but there is still parts of that technology that are 15 years old and so we're using AI to really help us.
(07:17):
We don't have the folk around the team anymore that knew that code base, modernize that code base in our target architecture in a more cloud native format and accelerate that work so that we can get done with that. And modernization is difficult and can go wrong. We operate within cost cap with less resources than typically you might see in a kind of digital business. This we rely partners as well in that space and we've really doubled down on AI to help with that modernization. It's now riping through all of our engineering, so not just modernization but new future development and our engineers have really embraced it, really led toward it. We now have our QAs and business analysts and product managers and others all leaning into this space as well than looking to, well, how do we use this going forward? But it hasn't been easy. It's not like we've just flipped the spread channel and it's worked for modernization. It's been really tough to get it to work, but what we are now achieving is phenomenal.
Jon Herstein (08:14):
So on the engineering side it's about efficiency, modernization and really taking advantage of these capabilities to get more done faster. And to be honest, I'm hearing this theme a lot in the nonprofit world given how squeezed everyone is from a resourcing perspective to be able to take advantage of these tools is really pretty powerful. But there's also some fun stuff that you've done including using generative AI for the mustache photo generator. So can you talk a little bit about that and what the impact of something like that has been?
Graham Link (08:44):
So they actually had two, so last year was the first time that came out and that was led out by UK and European teams and really got a ground grassroots swell around that. It was about playing with the fun side of ai. For those that are trying to grow MO or can't grow Mo, it was able to take a photo rock, something pretty iconic and cool on their base and share it with socials and share awareness as by vendor. And this year we've gone at it again this time with another partner of ours. So you'll start to see promotion of that coming through. And I have to say this year's ones that are really awesome, I was trialing it with my kids and my wife over the weekend and it generates some pretty awesome modes.
Jon Herstein (09:25):
So that's on the fun side. And then you've done something actually similar, something that we did at Box, which is centralizing a lot of your support documentation toolbox hub and then promoting that to users as a way for people to get more support in a self-service model. Can you talk a little bit about that, what maybe genesis of that was and what the benefits have been?
Graham Link (09:43):
Yeah, absolutely. So I'm talking about box hubs here and we jumped on that pretty soon as they'd come out pretty quickly. Quite exciting for us. So we have a lot of unstructured data at proliferate throughout the business and obviously within technology and support as well. And so efficiency played for us in using the box hub was being able to connect up all of that unstructured data that is in all sorts of places, really kind of hub over the top of it and expose it to our internal teams. So if somebody needs to acquire a new laptop or needs to understand how they can acquire some new software or get access to a platform, they can ask the hub and the hub pretty much we'll find it for 'em. We've done a lot of work over years gone by of documented all our procedures, our information's just people struggle to find the information. No one's going to go clicking through folders if it's on our intranet, we have something called Stash. If it's on there, they might find it. I'd say they more often they wouldn't, but the bops hub now is linked straight through there so they can go straight to the hub, do that search, find something relevant, and if it's not relevant then they can all ramp create ticket in that fly through to us.
Jon Herstein (10:48):
So you still have the option of someone still logging a case and getting support that way, but this is probably hitting a lot of that off. From an efficacy perspective, what are you measuring? Is it user experience, is it the number of support cases you get internally or how do you know? So it sounds like a better experience, but how do you know this is having impact?
Graham Link (11:06):
So two things. So we test internal engagement with our employees and understand how well the service desk is supporting the internal base and that's across the globe. We also look at all our tickets into equality and so volume and type of issues that are coming through, we can see the trends within there. So largely as with many organizations, there's a lot of self-serve opportunity within that type of ticket flow and so that's what we measure. We see more and more of that going to the box. I'd say we're still on a learning journey, our organization as well on even getting them to the starting line of self-help. And how do you get to the box hub in the first instance? It's November and we are very much a pupil thirst organization, so there's always the temptation, but in reality we work in different time zones and there are people in corners of the globe looking for help that would be 3:00 AM my teams support here and can now sate that. So it's been really powerful. We are using it in other contexts as well. I think the hub architecture that kind of sits on top of your unstructured data is a really powerful concept and what I like, we didn't have to move data around and files around, et cetera. We could just read it around.
Jon Herstein (12:11):
Yeah, we think that's a huge benefit that you don't have additional kind of content spr all that could happen. Nice angle. We also need to make it accessible over here in this place or that place that you just sort of point to it and it works. Are you already thinking about other uses for that capability for content management use cases or knowledge sharing or anything else?
Graham Link (12:29):
Yeah, absolutely. I mean I think our box data is around 40 terabytes. It's not huge for some olds, but it's pretty big for us and when I look at our operating model and how November goes about let's say health thing, our investments in that space, as you can imagine, organization of 20 odd Jewish, we've had a lot of people come and join, do some time with us and then move on. And I think there is a real risk in that type of environment where a lot of knowledge gets lost in the individuals as you move through. So over 10 years you might have seen three or four people come into a specific domain or a role really unlikely that a person in that role today knows everything that the person 10 use had known. In some cases though, that's a missed opportunity and November's been doing amazing work for a long time.
(13:16):
So how do we make it more accessible to all our employees? How do we remember this organizational information, make it part of our DNA so that if you join tomorrow, John, you could be up to speed with the last 20 years of November and I think you just box running for that's shifting or so years. There's a lot of information in there on how do we make it available. Again, you're never going to go through all full structures and be hundreds and hundreds C, you've got to surface it. So using it in that context, so for instance, having a specialized prostate cancer hub that sits across all our information in prostate cancer, that then also allows us to not just get to the files but to digest the information that's within there and surface that up is the incredible path.
Jon Herstein (13:56):
Yeah, I think that's a great use case for onboarding new employees. We've also done something similar with some of our key roles where we've created onboarding hubs for new employees because to your point, there's a lot of built up institutional knowledge that new person that just doesn't have access to in their brain, but if you can make it easy for them to access it, previous decks from our all hands meetings and strategy decks and dashboards and reports that they wouldn't have had access to previously and now they could just go ask those questions. We found it to be a pretty powerful thing. So I love that use case. How about things like localization and translation given the 20 plus countries of presence that you have, have you started thinking about leveraging AI around content for that?
Graham Link (14:35):
Not so much on the unstructured data, the majority of our content would be in English. It just happens that the majority of the countries we operate in, that is first language that we have within there, but absolutely context of our website and how we turn up around the globe. Then yeah, a lot of it isn't in English and how we use ai. We're exploring opportunities at the moment in that space as to how we can better translate and expose our content, other languages as well as into other AI models and removing growing SEO and space and into these newer concepts around consuming your information in a q and a type format into the R of the world out there. So definitely in that space
Jon Herstein (15:16):
There's some really, really cool use cases and I can see just a lot of business benefit coming from those and I know you'll continue to explore more. I want to shift gears a little bit and talk a bit about leadership, how you drive some of this change, driving change in culture and so forth. So maybe I'll start with Movember's not a very old organization, I just had a podcast recording with an organization that's 150 years old, so they've got a lot of built up legacy. You've got the benefit of not having that, but I'm curious how has the organization built a culture that really embraces transformation and specifically in this case digital transformation
Graham Link (15:52):
In past because it needed to. And so when I joined November and looked at it, it looked a lot more like a startup scale up structure within there. And what I mean by that is people will laugh for a real purpose, all want to be that number one really prepared to wear multiple hats. You've got to be able to jump out of your day role and pick up other ORs and support the team in completing whatever that program or initiatives we're going. And so we've always been quite bold. I think my vendor and willing to challenge and our values speak to that around fun and change agent and accountable responsible and it's really embedded in I think who we hire and how we think and how we operate. We're our biggest star critic and we're always pushing to be better and really driving hard in that space. But the team I've got on the ground within technology and the broader mode vendor team, I think that because they're up for the challenge and the status quo doesn't work in November, the rule changes too quickly around us and we need to adopted that as well.
Jon Herstein (16:49):
I love that you have fun as a core value and I would say dealing with technology is not always fun, although these days it is pretty fun dealing with what's new in the innovation, but how do you sort of maintain that fun and core value but also deal with the reality of tech and the challenges that come along with that?
Graham Link (17:05):
Yeah, grass is always green on the other side is that sometimes when you're inside you feel like you're in the fry hang. I think a lot of that comes back to relationships and how we build a culture around people and value people within November fighting back in my early career and what learned from that was some of the hardest toughest projects I ever worked on were some of the most fun and the best because people we had, I think November wouldn't exist if it wasn't for November supporters and the operators of November and people work within the organization and so we put that up as a high priority for us. We work in a hybrid situation now post COVID probably need to say stop saying that now because it's just the new reality. But we work in this hybrid filtration which makes it more challenging to connect and we have distribution as well, so across Australia where a lot of the tech team are but also in some other markets. So we're really deliberate to make sure that we have scheduled get togethers, start meetings, always make sure there's a five minute social connection within that. We encourage, allow our employees to come up with sort wacky fun things to do in our ideas and walking into our office is quite interesting on Sundays it is genuinely a lot of arts but it's tough as well. It's not just large rules that the work is hard, work is purposeful and rewarding and the people sign
Jon Herstein (18:19):
Well and I think when you've got a mission-driven organization like you do and you've got I think the benefit of, again being a relatively young organization, I suspect your culture sort of naturally embraces change, but change is often hard for people. So I'm just sort of curious too, how do you think about driving change in the organization? Is that easy? Does it come naturally or do you have to be very deliberate about things like okay, AI's coming along, we want to adopt these technologies. How do we actually get people to do that? What do you do to make that happen?
Graham Link (18:47):
There's a couple things to that. We have to be very deliberate. I think change is really hard side and part of that is because of the pace that we operate in where our focus is. If we think about the fundraising side, we've got one big opportunity. You have to do fundraising. For instance, we're in day 5 20, 25 November, but we are already planning and scoping 2026 and so when change comes in and disrupts those processes, that's really quite hard. And particularly if you're talking about AI and use of new technology, new ways of thinking, even approaching problem, you sort have to make that space for it within the organization and within the work that we're doing to learn and stop that technologies make a lot of mistakes and get the value in it. So for us we have to be really deliberate. We over communicate a lot of things and some of that probably results in too much information at times, but usually the feedback is, oh, we didn't know about this or if I knew that was going on or that was available. And so over communicating and we try and spread it around the organization as well.
Jon Herstein (19:46):
Is there anything in particular you found with AI specifically that you've had to do differently from a change management perspective? Any kind of different messaging or sentiment that you need to communicate?
Graham Link (19:55):
Definitely ai. So a couple of things. First of all, got more stakeholders involved, got a lot of employees, all ages as well. So you've mentioned earlier we haven't been around that long. We've got people straight out of university through to have completed a career and now working in a more purposeful space. We've got employees coming in with prior experience or expectations using this terminology or maybe some concerns about this technology. I'm not sure what it means for them so we need to manage that Greg, but we've got board as well and execs who are wanting to know how are we using this technology, what is our plan and our strategy around it. A year ago I had planned John to develop our AI and data strategy and take quite a traditional approach to doing that. We'd spend a couple of months really doing discovery around that, identify where we want to be, what's the plan and I quickly realized this space is moving so quickly that that framework wasn't going to work for us.
(20:47):
So we adopted a much more flexible framework and we went to one of the core tenants in November of cross-functional working. And so this guy addresses how we do part of the change. We put in clay, a senior cross-functional working group for AI in inside November representing all the different parts, so be it from legal, the institute, technology, fundraising, marketing, et cetera. And as we've convened that group and we effectively built a AI enablement framework with that group of how we're rolling out into the organization. So having a framework is towards allowing our employees to experiment safely where it makes sense but also we using AI with sensitive data, let's say in a health context and from appropriate guardrails and governance is in there. So getting that cross-functional working group is really important. That also acts as a connection points through our employee base, through their own departments and so someone in fundraising can go to fundraising representative on there and have a conversation about how did they get some ground and take it forward. And so I chair that with the CEO and I think having that CE executive support from the top down of potentially makes us better. We want to lean into this but in the right ways.
Jon Herstein (22:00):
That's a really interesting structure and it sounds like every organization it feels like to me has built or expanded their governance process to include aspects of how do we think about AI and bringing AI into the business. This sounds a little bit less like governance and more about ideation, brainstorming, visibility, that sort of thing. Is there also a touch of governance or what's the full scope of that
Graham Link (22:21):
First session with ram? Our working group was a mindset session and it was all around what is the potential of ai, how are others using it? And really try and level up the understanding of the group of this is where we're at today. Even if you've been doing some stuff or seeing some stuff that AI may be at home playing with it a little bit here and there, this is what's actually possible right now and what we're seeing in the industry, both nonprofit but also commercial, what's going on that, so we did the level up and we did some opportunity identifications. We very quickly then switched on to get this stuff through. We're going to need governance, so try and make more governance liked that's fit for purpose. The internal tassle we have John on the governances, we already have governance for technology and AI is one case of existing governance, but we've had to make it so clear by breaking out of that and just extending existing policies we have and existing reporting lines and frauds or reviews for instance, like doing a supplier assurance assessment, which we would do for any new supply being very deliberate that if there's an AI platform coming in, it has to go through this process so that those in the organization are looking, implement their AI idea, understand this is how we do it and they may not typically have interrupted with technology that much in the past, but in this context with AI that may well be a lot more frequently.
(23:35):
So yeah, we've had to bring both sides to it. And then so the working group sit across both of those. We've also built an enablement team under that. Enablement actually is hands-on helping projects within cyber November bring their AI initiatives to life.
Jon Herstein (23:49):
This is a really fantastic structure because what I've seen a lot of these organizations wind up with the governance committee that sort of become the thing that slows things down and it feels like this has at least repurposes. One is to come up with these ideas and socialize them. Two is to impose governance as appropriate and then the third is the enablement piece that you talked about. So I would imagine people look at that group as something that's there to help as opposed to I'm trying to do something with AI and I'm being told no.
Graham Link (24:16):
Yeah, absolutely. The enablement framework that we have allows people to understand the risk of what they're doing and then based on the risk category though, how they need to interact with governance, we also have to register for ar, so you can check that and go, okay, well let's say it's box using box and we're good to use it for this kind of data as well. We can use box for highly confidential data and then maybe they want to build an AI workflow in context of box and so that'd be something that we'd have in our register. They may look at that and go actually already been done. They'll have a step to record the new use case, but effectively it's a fast process for that to get to implementation. We give the others John that require ethics approval. They're slow by nature and so they have a better conservatism and spending too much money earlier and unless be noticed ideas and that flow through, but you've got to do enough in order to get some of those hops through but that could happen whether it is AI or not, they're just part how we carry in technology in health.
Jon Herstein (25:10):
Yeah, that makes perfect sense. You mentioned risk a couple times in the context of governance and I wanted to kind of point you to a study that was done pretty recently, university of Melbourne and KPMG. The study was called trust attitudes and the use of Artificial Intelligence and they found that 48% of employees admit to uploading sensitive company information into public AI tools in over half present AI generate content as their own, which is pretty shocking in a business context. I think we all worry about that in an academic setting but in a business context it's pretty shocking. I guess I'm sort of wondering for you as a CTO that is embracing ai, how are you balancing empowering the team and encouraging them to use tools and get the benefit of the tools with making sure that the usage is responsible and ethical and
Graham Link (25:56):
Secure and private? He went about prior chats. John, you had a phrase which I've used since and I love which is bring your own ai. We used to think about bring your own device, we still think about bring your own device and we've got technology solutions for that now we see bring your own ai. I don't think it manifested in other technologies as much. No one was necessarily bringing their own emails and that report that you mentioned, I read that too and those stats are global across a lot a lot of markets and what's going on within there. So there are obviously technical controls and things that we can do within existing policy and we do do some of that. So we deliberately do stop certain types of information being uploaded into unauthorized platforms and so having clear education to our employees on what's okay, what's not okay, that's been leaning into and that's actually a big part of that enablement framework on what you can do.
(26:48):
We're very quickly moving to help the right tools in the hands and we've been running a pilot on some journey of AI platforms for the broader employee base, which has been fantastic and a lot of positive team upside to that. So I think getting the right tools, people will go towards a space where there's a gap and people see a lot of value in this technology so plug in that gap but it comes with a cost and so for us as well, it's not being just we want to just throw money at there, we wanted to understand the value that comes from it and then how we run a dividend investment model so that our AI is funded by productivity within your current,
Jon Herstein (27:27):
But it sounds like a combination of policy guidelines for folks but then also providing them the tools you have validated that has gone through your committee and that you feel comfortable with. And if you're doing all three of those things then I think you put your employees in a better position to do the right thing.
Graham Link (27:45):
Absolutely. They all come into it and I think the AI part comes at different angles as well. We recently just did cybersecurity training again against the whole employee base, but they were AI based cybersecurity training pieces talking about the use of these tools but also how these tools now being used for instance to conduct a real sophisticated phishing attack and it's about leveling up the understanding of the organization of what AI can do and what to be aware of so they're hearing it in different ways. There's so much potential with that technology for the workforce I think it is around having the right tools and the right support. So you're have to train your employees and develop them and don't assume that most of them can just do this, some of them can, but there are others that have never touched it and the other half of that stat is then ones who have to be brought along and shout out to how to use it effectively.
(28:35):
I also see a lot of people, John using AI and even with the pilot we just ran recently for one of the major LLM players out there, people are using it but not necessarily effectively. We had to run internal training on this is how you really maximize this platform and working in ways that may not have worked before and the way that you prompt and everyone's had to become a bigger prompt engineer and understand about building project knowledge base and so that when they ask questions and they all seek to support a document generation for instance, it can do a really effective job.
Jon Herstein (29:05):
We've done something similar. We've done our internal AI certification both kind of broadly around what did ai, how to use it, that sort of thing, but also our own product box AI and we'll be rolling out to customers as well as a way for people to get up to speed. You mentioned people and obviously there's a huge people component to all of this and I'm sort of curious, we talked a little bit about roles and how are you seeing the evolution of roles, things like knowledge managers, I don't know if you've had knowledge managers in your organization, if you do or you have sort of adjacent role to that, how are you seeing them change and then we'll get into a little bit more about skills development and that sort of thing. Maybe we'll start with that. Just role changes.
Graham Link (29:41):
I think we're quite early on seeing the roles change. We recognize that this technology can augment a lot of roles if not most roles within the organization. I would say to date that most for the pilots we've run and the initiatives that we've been standing up, people are really l into it and embracing gone, this is going to make much more impactful, it's going to allow me to perform my job better. We're always a capacity constrained organization. There's always more work to do that we can get through in all the s and so the potential of using this to be more accurate for us time get to the pump quicker. I know that's something we'll use it to current digest, really large document tech theme and really get the kind of rain points out of there that the might take. So an hour and I can do that in minute.
(30:23):
It's such a simple use case but there's much more interesting ones I think that comes through the area. I've seen the change the most and maybe this is a sign that others as well as we've gone on journey is in that engineering use case by far the most sophisticated AR use case we have today in November, we've really flipped the model on how to code and how to engineer and in a traditional PITA team and we look at the construct of who's in there. We're now realizing that first of all we need to produce senior engineers into this problem space to really develop our own gentech AI workflows around these coding modules. So there's a modernization piece there that there's new development pieces as well, but then when you step out you go, well you're very much able to do that in systems thinking and acting like a full stack developer engineer, an architect you won't go, but what else do need?
(31:13):
How do we become more effective? We start see the emergence of new roles in the team and we're not there yet in implementing them because people are wearing multiple hats, but our mind goes to next year's budget and how we're going to set our engineering teams up and we're probably going to have some new roles that we didn't have before for it's prompt engineer, how does it prompt engineer Now it might want to be an existing software engineer who has the role, a prompt engineer for the team as well, but today our lead architect is our, I'm going to be our prompt engineer. He's the one that can get the magic out it and I think it is challenging. I think the roles will change in engineering and based on the new way that we're working, but it's say and in general our team have really lean into it.
(31:52):
I think we had a few questions early on around will I still enjoy this if I do so far very positive and people are enjoying this new world because I think in the engineering use case, most s they're problem solvers. You like to solve the problems and there's this learning uptick that might be a bit confronting initially then maybe someone haven't had to do for a while. But once you've done that and as you go through that process that the problems we can now solve are greater and more complex and I think that there's a lot of enjoyment there. We do things like our documentation is automatically updated John, so as code changes it's automatically refreshing all the documentation surrounding it. That's a job I've never had an engineer telling me they love doing but it's something that the AI does really well and very happy about that. The testing side quality assurance is phenomenal. The coverage we've been able to get in a very short timeframe through this is mind blowing compared to what we used to be able to do.
Jon Herstein (32:50):
I think this is a really key point. It's taking some of the tests that frankly a lot of people don't enjoy doing anyway and allows 'em to focus on the more creative, more strategic aspects of the role. At least that's what we're seeing and even if there's some initial skepticism, it seems to fade pretty quickly when the light sort of clicks on.
Graham Link (33:04):
Yeah, we talked a lot to our team early days John, about how we hire an individual based on their experience, their skills, kind of personality characteristics and so on fit for the team and then we put someone in a role that I would say 80% of what they do, any person could do, any engineer could do that and really there's only a really small portion of the time in that role that their uniqueness is brought to the role. I see that in lots of different roles and one of the flips here is that with the AI supporting more of that tedious manual stuff that you don't need to do as much of anymore, you bring more of you, you bring more of your unique position. What was your unique experiences and how you can have more of that involved and that might be more thinking time.
Jon Herstein (33:52):
How do you think it's affected you as a leader when you think about both using these tools as a leader but then also leading the organization through using these tools? Do you have a sense of how you've changed?
Graham Link (34:02):
Probably initially super excited about the idea of this. Also a little bit pessimistic is to, well let's see this play out. Lot new vendors is emerging, seeing these things pop up before and they're not quite hit the mark and where will they go to be absolutely blown away with the pace of this technology as it's coming to the market and what it can do. It was always on the inside and wanting to see this come to life but maybe a little bit more reserved in expectation of what's here to having to jump in a lot more on if we think we can wait two or three years to see this technology play through and how to use it, we more just be playing catch up. So needing to move to a much faster base and understanding of the technology. So I get hands on with it. It's really now about getting the whole organization to adopt though and to leverage this where it makes sense.
Jon Herstein (34:51):
Well, and I imagine your peers in leadership at November probably are also at varying points along that continuum of being pessimistic or not pessimistic but maybe a little skeptical of what's possible or how quickly will happen and others maybe fully leaning in. Have you had to bring the team along, not just the team that works for you but also the leadership team along and also properly manage expectations about how much of an effect this will have?
Graham Link (35:15):
I'm probably able to pull it back a little bit at times now, so I think your teams always be really open to the idea of this and very keen to see how it come through. There's definitely some reservations as to are parts of our organization ready to embrace this? Could we pull this into one of our core functions, say finance or marketing, really adopt it in a wholesale way and there's a bit of a journey and learning to go on which is totally natural. I think the thing that changes is that you're just seeing the use cases in the industry now at copper every day and just more and more information slowing out there and so whether it's board directors or just being immersed in the industry, what you hearing seeing it is causing a lot of interest for this but recognizing it's change and November's really good at adding stuff to do and adding more things to do and they're not very good at taking off harder things and slowing down at times. So that's just the nature of who we are and sort of wanting to lean into mental health and really drive change. And so there's always like that.
Jon Herstein (36:17):
I would imagine part of the message is that this technology, these cable, and it's not just ai but technology generally is really an enabler of the mission and it's not just something to do for fun but it actually helps you get to where you're trying to get to faster, more effectively with maybe fewer of the same resources. Has that been kind of core to your message to other stakeholders?
Graham Link (36:37):
Absolutely. So when we talk about that, why using ai, I mentioned it, it comes back to because it can make us better and does that in a number of different ways and I guess when we think about AI as well, we do put it into three camps and we're still working through the answer this we haven't done necessarily know what's going to go into the three buckets on the cap. So one is where do we want to be leader in ai? Where does it make sense for us to break ground and really lean in and for instance, the modernization we're dealing with cloud there with some of our other partners as quite new counting and that made a lot of sense for us to take that opportunity on. There's a second piece which is around where we are fast follower, we can allow others to carve the way, but once it's available to us we'll adopt it because there's huge efficiency price for us and it makes more sense.
(37:21):
And then there's a third category which is where will we not use ai? Where will we deliberately remain human? And all of that comes into our communication and interaction with those. No, it's a movement go on people and we very much need to hold and preserve what is authentic and true about our organization. However, it's easy in the individual case that to see it that way is another challenge where we think on a global scale, so while we're reaching millions of people or in a very short week of time, for instance over campaign technology, we couldn't do it without technology and AI as a world play as we go along on that journey. More in the use case of I think making sure that when people are participating in November, so fundraisers or donors that are a part of this and interested in men's health, how do we get you, we've got so rich information about men's health, how do we make it relevant to you? That's one way AI can be really powerful coming through and giving you really quality information and maybe it doesn't connect with you, but that's easy then for the AI to determine that was a light and here's the other path to go down. It could have been more horse street or heavier and slower in path.
Jon Herstein (38:30):
Well I love that idea of just in terms of maintaining authenticity, being really deliberate about areas of you say we are not going to use AI for this thing because it has to be from us and it has to be genuine and authentic. I think that's a great way to think about things. For me it's something like if I'm sending an email to my broad team or to one person or to a customer, I'm not going to use AI to write that, right? I'm just not. If I'm doing a mass email for marketing campaign, well maybe that is appropriate to do that as long as it conforms to your voice and your brand and everything else. So being very thoughtful about it I think is very important.
Graham Link (39:05):
Yeah, so we've said that I've been our AI framework framework and around the part of that is we've got principles and so basically talk to those points and we give examples of we can use it here but we wouldn't use it here, but exactly it's that authentic voice drawn, right? Even though the AI can generate a very authentic voice, if it's not authentic, then does it hold the same value? And probably for us it doesn't staged, but behind the scenes and what it can do for sale and volume, it's pretty exciting.
Jon Herstein (39:34):
Well, and like you said, you're not setting rules that are going to last forever. We all know this is moving quickly and so what you might say is off limits today, maybe six months from an hour or a year from now, you'd say actually it's a great tool for that use case. We're still, I think in the early innings of all this do have, are you willing to share an early failure or a pilot that you tried that didn't work out and you're like, okay, we're going to move on from that. Particularly it's not so much about the thing itself, but more lessons taken away from that that maybe would be useful for others.
Graham Link (40:00):
I said we're pretty early on in our general around risks, at least with the generative AI capabilities and the generative AI componentry that's spill in there. I would say we are, so we're about seven or eight months into that engineering. We kept ready shipping the model on how we develop and we're already three tool sets in. We've already shifted multiple times, so maybe less about that was wrong and a mistake, but more that it's new so quickly around this we needed to be nimble to move as the end issue's moving and there's things that we built six months ago that are now just features of technology IVs for instance, that are there that remind we just put all this efficacy's available. So there is this question mark here on how long do you wait and watch what's happening for jumping in, getting stuck into that.
(40:46):
We probably thought it would be easier than it was to be honest. It'd been much higher to build out that engineering organization workflow and that ag flow let's be in there and we got caught up in it getting caught in itself and constantly trying to fix defects and then introducing new ones. And there's a whole lot of techniques and patterns now for those things, but they're kind of emerging as we were emerging and laying the groundwork for that. And we're still only about a third of the journey through being in space on the whole, it's still super positive hyper failures, those issues are all just part of the learning cycle and I don't think it's possible to go and take this type of technology on and implement it and get it right first time in the type of use case that we're adopting and what we would go through, which is why we would be a bit more reserved say using inner health impact, like delivering something that was facing men but had an AI in it. We would want to go through a lot more checks and violences before that. So different risk tolerances for internal versus external and then different data types as well.
Jon Herstein (41:47):
That makes perfect sense. And we're not dealing with health information ourselves, but we thought a lot about enabling our, for example, our support organization to use AI to get better answers to customers. That's very different than letting customers get their own answers with AI and running the risk that we give them some bad guidance so the bar is higher. So I want to talk a little bit about as we work towards wrapping up here, three areas that I think about a lot in the world of customer success are the idea of delivering value, culture and change and the experience that we provide. And so I want to ask you a couple of questions on those three dimensions super quickly be relatively quick. We think about value and delivering value. How do you define and measure the value that AI or any new technology arguably is bringing to Movember?
Graham Link (42:32):
I think you'd get a pretty good test in talking to the individuals involved in this and that's not how I would fully measure NR on this, but given how LA some it is now early days we are, you get really great feedback on the team is to are you enjoying this? Is it improving things? Well why is it improving? So we've done, there's just the conversations to be had and we've done with the pilot we recently did for our gen AI platform, we've done studies on it and we've actually tested and only got quite ER around what are the efficiency gains, what do you like? And so we're building this kind of pulse survey around how is this working for you? Is it improving quality of your role? Are you enjoying your role more? Has it taken away things that were distracting you from the main point of value that you had?
(43:18):
So we get a lot of pieces of feedback, quality of and quantitative in that. When I think about ROI though in the bigger investment parts that we're doing, not to necessarily go back to that modernization one, but that and in general now we can see real value return on that. What we're able to do with AI would've taken us multiple euros and a whole lot more people to achieve full scope. So there's a real dollar return on those pieces. I'm a little low to lead with the dollar return on any of this at the moment because I'm more interested because it's evolving so quickly of this idea that AR can help us be better. And so how does that manifest itself in an organization way? Does it make sense? But in order to get the investment up and to have the funds available to purchase those tools and have a small team to support it, there has to be some revenue impact to that upside on that or have some optimization that we do look for
Jon Herstein (44:11):
Probably it sounds like a little early to get really hard ROI, but you know that you're deriving value because of the user experience or how quickly people are to get things done and you're beginning to measure those things.
Graham Link (44:20):
We've got a batch of people that want access to this technology that we've not yet opened up, but it will come probably the about there is the quality of the output and how actually is the quality changing. I can definitely see it in the technical use cases that the quality assurance side to what we're did massively improved. I'd be a little more interested to see how that might help in a more general setting, say in an operation setting within the team and what that's meant. So that would be designed, I think those kind of measures are bit science specifically around the initiatives we do and I should have mentioned, but it's part of us having a register of use cases that are approved and a framework to enable framework. Part of that framework is to measure, so it is to come back at six months, however months as to how, what is this achieve? Is it working as intended, wherever it works and have that open dialogue.
Jon Herstein (45:11):
That's fantastic. And again, having that framework as a guide for all this, I think it's got to be very, very valuable to you. And I would imagine that you evolve the framework as you're learning too. It's not a static thing. Can you say a bit about user experience? You did mention one of the measures of success is are you getting other people who don't have access to these things, clamoring for them? How does that show up from a user experience perspective?
Graham Link (45:32):
So protect the gen AI pilot. We've been running with one of the big platform providers that's out there. It is very much a hop out siloed experience, so you jump into that tool and you'd be using that. The experience itself was pretty good. We had to do some up stilling and development of people in that pilot, so then you had to use it. But in that timeframe, what we've now seen is the integration mLab plays out inbox as well. And so that platform is now connected up to one of our other systems where it's like Outlook and Slack and CRM and other parts. So now this tool has become a lot more connected and able to query these different data silos that were there and so people aren't necessarily hopping out as much. It's also been exposed through other channels so people can use it in different ways and kind of still seeing that paint out. But I'm just seeing this kind of connectedness and glue really rapidly come into the ecosystem. It feels like every week or two there's a new theme there that we can have a look at. And Ridge,
Jon Herstein (46:31):
I think you're completely right about the interconnectedness and I think what we're going to see is that integrations are going to become a lot less rigid code and API driven today, it's going to be agent to agent tomorrow, and that's going to, I think make it easier to connect things up in a way that feels very natural to users. But we'll see how that evolves. You can already have one simple product thing and with the box Slack integration, you can have access to box content and Slack and ask AI right within your Slack thread as just one really simple example of just do the work there, get access to your content and get the benefit of AI in there.
Graham Link (47:04):
Yeah, I love Bozi spaces and the less hops, but just that utility including fact that've been able to do it. And we're thinking about that too, John, we're thinking for next year's campaign, how do we stake MCP servers or other exposures on our fundraising tech? It's hard to predict for how it's going to shift, but in 12 months time people might want to register through their ai. Some might then, or maybe they want to donate through their hour to November. And so we needed to start to jump on that as well and start to explore how might we make that possible in that ecosystem. Yeah, it'll be here really quickly on the current pace to change. Super interesting, exciting as well and kind of breaking ground in a whole new era of fundraising that and maybe we haven't done before. So
Jon Herstein (47:47):
You mentioned MCP and I was actually going to ask you what emerging trends are you excited about that may have the most impact on Movember, so MCP potentially to help with fundraising, make it really easy for folks to do that. Any other ideas that you're kicking around or excited about.
Graham Link (48:01):
I think sticking to that talent around authentic messaging where we think about November's reach and our fundraising platform around the globe, there's a really interesting AI use case there to maybe tailor the content that's relevant for certain markets and individuals down to a micro level so that if you are living in a given city, how do we make sure that by vendor resonates with you in work that we are doing that is absolutely relevant for you, that maybe is less relevant to someone on the other side of the planet in a small town. So I think there's some potential there for us to cut data and localization and geography in different ways using some of the AI smarts to help us get there whilst also staying authentic though using our actual words and our data there. So I think we'll see some of that play out as we go forward.
(48:52):
I think there's a big question mark on November's role in health and AI and what we'll do here, and we've got some partners at the moment that are playing in AI and there's an amazing initiative in UK at the moment around cancer, and so there's a platform that we're launching in 2026 then to provide for free support to men go through prostate cancer and breast cancer in that construct. We are an information provider. We have a really also low site culturally north, which has a library information around survivorship in cancer and managing your journey and connecting really good quality Asia that's not out there today. And so we support that AI through information, but there's a question mark on is our play information or is our play AI as well and time in else unanswered John? But that where we cloud line to going forward is to what is our role as what are the role of others in that ecosystem?
Jon Herstein (49:46):
Yeah, and who do you partner with and who does what? I mean there's so much opportunity and so much like you said, unknown right now about how it will all play out. Any predictions
Graham Link (49:55):
Then reality is that some of 'em are quite advanced, but a lot of us are on catchup and leaning in heavy, narrow, but racing against something that's for thrust to virus as well. Sorry, I'm not sure where all are, but I think it'll be pretty positive to the way we go. We'll lay some bets in how residents, so over the coming years in the space, that will be super interesting. I'll see them, see them fly up.
Jon Herstein (50:17):
Do you have any tips for a fellow C-I-O-C-T-O who's navigating this journey? Maybe not quite as far along as you all are? Actionable tips, any guidance you would offer? Mistakes to avoid, things to think about?
Graham Link (50:33):
Coming back to the early part of our conversation, having in place a good framework, but for managing ai, and there's a lot of, are probably working with the big tech players, the cloud platform providers, are there a lot of insight as well into how others were implementing AI and their frameworks and so I'll be leveraging the partner ecosystem, like we have to bring current thinking and best practices to that. I'd start managing a board now as well as the employee base, so board executives, people leaders, and the core workforce depending on the nature of the organization as well. Obviously I'd be on the front foot with managing up all levels and getting the right governance in play as well, and not necessarily over engineering as thing. The idea of a three year strategy for me in AI just doesn't connect as well as say a three year technical architecture vision does.
(51:22):
There's a lot clearer as to how, and you could be mad those other, but with the AI maybe so quickly to suspect after year one you'd rewrite it. The other tip, and I give this tip not just to my level is everyone in the team is you've got to get, it's very easy to get caught inside the organization with the pressures and priorities that would be in that organization and we all fall into that trap at times. I do most of my learning, John, just getting out and talking to other organizations and other CIOs, CTOs and leaders and how they're doing this and that's where most of my learning happens. Do more of that and share the ideas. And some of it for me is informal. Like we have coffee catchups, the small groups in primary Melbourne in Australia. We've got a community there, but I go along to various meetups and conferences and get involved and usually my biggest learning opportunity is not, I'm not so much interested in the kind of tech partner and what they're promoting necessarily and what's going on there. I'm much more interested in hearing from others and what they've done. Got to do that this year in the us in the non crop space and add some really amazing examples of AI being implemented and really starting to help us think about how might we approach this.
Jon Herstein (52:32):
Well, that idea of sharing and learning from others outside of your organization is really key. And actually one of the main reasons we're even doing the podcast is to give folks like you an opportunity to share what you've learned and help others avoid maybe some of the early mistakes and learn from each other. So I really, really appreciate you Vember, the time you've taken here today, but also the long partnership we've had together and really for sharing the incredible journey that you've had in bringing this AI innovation to November and helping advance the mission. Appreciate your time. I think your insights on balancing the technological advancement with human-centric values and the practical examples that you've shared on implementation and the nonprofit sector have just been really, really valuable and I hope people get a lot from it to folks who are out there listening. I just want to say we are releasing this in November.
(53:22):
In November. It is the signature month of action for November, so if you can get involved, if you can donate, please, please do. You can get involved with one of the many, many groups that are out there. We appreciate it and I'm sure Graham does as well. And we really appreciate you taking the time to listen, to learn, and we thank you for joining us on the AI first podcast. Until next time, keep pushing the boundaries of what's possible with AI and consider growing a MO for a good cause. Thanks everyone. Thanks for tuning into the AI first podcast, where we go beyond the buzz and into the real conversations shaping the future of work. If today's discussion helped you rethink how your organization can lead with ai, be sure to subscribe and share this episode with fellow tech leaders. Until next time, keep challenging assumptions, stay curious and lead boldly into the AI first era.