Exploring the Potential of Microsoft Copilot | Interview with Tom Arbuthnot, Co-Founder of Empowering.Cloud

Ryan Purvis 23:50:43
Hello and welcome to the digital workspace works Podcast. I'm Ryan Purvis, your host supported by producer Heather Bicknell. In this series, you'll hear stories and opinions from experts in the field story from the frontlines. The problems they face, how they solve them. The areas they're focused on from technology, people and processes to the approaches they took that will help you to get to grips with a digital workspace inner workings.

Welcome, Tom, back to the digital workspace works podcast you want introduce everyone to yourself and your new venture?

Tom Arbuthnot 23:51:17
Yeah, thanks, Ryan. Thanks for having me on, again, really excited to be on and catch up. So I think we spoke about this previously, briefly, but I'm running a company called empowering cloud. We're focused on Microsoft Teams modern work, and it's spanning a whole bunch of community work, and then some premium offers for Microsoft partners and for ISVs, OEMs, who want to get deeper into the space basically.

Ryan Purvis 23:51:39
All right. And we talked about the community driven approach, how you found that the sort of adoption of people coming to the community and, and your plans for next year?

Tom Arbuthnot 23:51:50
Yeah, that's super important to me. So the whole point of this was really to be community first, because I don't think that the speed of this cloud world you can't keep up on your own is a community effort. And I've always done a lot in the community. So at the moment, we have about three and a half 1000 people signed up, which is really, really good. It's still a fraction of kind of the stuff I do on LinkedIn, the stuff I do with my email list, but it's it's engagement. I'm really looking for that. But the big numbers are great for marketing. But the reality is, are we delivering value to people. And we've had 24 subject matter experts, MVPs MCMs, people from the industry, doing these 15 minute briefings, and they're all contributing those to the community. So that's a really big one for these people are giving up their time and expertise to share with the community. And that's the plan is to keep building that out. We're adding a forum Asti at the moment where we'll have more direct engagement between people as well, which we think is really good value. But yeah, early days, we're just coming up to a year. So for over a year in terms of where we started to where we are now I'm really pleased

Ryan Purvis 23:52:52
is that community doesn't have like a tearing thing like a free level and a paid level that or is your sponsors, paying for the

Tom Arbuthnot 23:52:59
Yeah, the kind of the model is 90%, sponsor back free content. So that's the thing, we really want to be like not just lip service to community, but actually deliver loads of value there. But there is a tier where Microsoft partners or vendors can pay us anywhere from about 500 to 1000 pounds a month, depending on headcount. To get Learning Path certificates office hours with me and the team ask their own private questions. So no hard sell there. But for those who want to take their entire team, for example, pre sales teams is a very common one, they want to get that speed, they want to know what's going on in large teams and they wanted to stay up to date, we can put them on learning path and the system will proactively push them video content via email, or via bot message all sort of 15 minute or less short, snappy content. And that's been really popular with partners who are struggling to keep up just just the same as we all are.

Ryan Purvis 23:53:47
Yeah,no, exactly. And you say system if you build something all up?

Tom Arbuthnot 23:53:51
Yeah, so I worked with a fellow MVP, Tom Morgan, who does all the clever stuff. So I can't take any credit for it. But yeah, we looked at various learning management systems. And we couldn't find anything that quite fit the bill of what we wanted. So we've built a platform on Azure, and we're using various component parts. So we use author, euro for authentication, using Wistia, for video hosting, using intercom, some of the communications. And we put all those pieces together. And that creates a platform where we have ultimate control over the user experience. So we can, for example, be very granular about who gets access to which content, we can see we have Power BI reporting built in so we can see who's engaged what they're engaged with. And that's really useful to say, Are we delivering value, but we also have research papers that we build in on Power BI, so things like teams, bar comparisons, or price comparisons. And we couldn't find anything that would do that kind of how many LMS is let you embed Power BI or none, because it's very nice use case. So So we ended up rolling around. It's been it's been hard work, you know, development is never as well. Anybody who does it knows that it's always always more than you think it is. But I'm really glad we own our destiny and as the platform.

Ryan Purvis 23:54:58
Yeah, that's important. I heard you say and we ended up building something ourselves now, just because there's lots of nice tools that do nice things, that you find that the integration between those different tools to each other using something like Zapier or whatever it is, will always be limited.

Tom Arbuthnot 23:55:14
Is Yeah, yeah. And you're you're you're in depending on your use case, or if you're doing a very generic call. A lot of these platforms are built for more b2c scenarios as well. So we want you know, high security with all zero we want office 365 login, we want LinkedIn login. No LMS is care about LinkedIn login, because there's not a use case for them. And the things like the Power BI and the Azure stuff, and now I'm super excited because we're starting to play with the AI engines in Azure and this is a way out but We have all this information from the subject matter experts. We have all the transcripts, we have all the content. In theory, we can run an AI engine over that, and have a dedicated dataset to start answering questions with a bot. So, early days, we're still playing with that. But I'm really excited about the potential of this evergreen not knowledge being more accessible to people.

Ryan Purvis 23:56:04
Yeah, it's that boats. What's that analogy? Or waters float on boats or high waters float on boats? It just feels like there's always something pushing everything up and up and up. And your barrier to entry to do something is getting lower and lower. Yeah, what's going on? You get drowned?

Tom Arbuthnot 23:56:22
Yeah, it's interesting, isn't it? Because like, everybody's everybody's potential is going up. everybody's expectations are going up as well. So you gotta keep up. Yeah, we couldn't have done I mean, anything like this. Even five years ago, in terms of the maturity of the cloud platforms were building on and the level the amount of development you can do I mean, Tom's, you know, leading all this from a development point of view, it would have been a team of people, but because we're leveraging cloud services and and loads of Azure staff, you know, you, you just plug it all together, we're probably doing a slightly service. But you know, it's, if you had to build all of this from absolute scratch, it would be a massive project.

Ryan Purvis 23:56:59
Yeah, for sure. For sure. So tell me a little bit before we started around copilot. And I was curious to get your thoughts on that and where it was going.

Tom Arbuthnot 23:57:07
Yeah, there's mad excitement around copilot at the moment, I'll see Microsoft are riding the AI hype train like absolute pros, in terms of everything is co pilot, the copilot and Windows copilot, then 365 is copilot and get hub copilot insecurity. I mean, it's always becoming a bit of a meme. Now in terms of everything has copilot, there's not, there's not been a lot of hands on outside of Microsoft, they have a couple of 100 customers field testing it. So in terms of hands on, like I've had hands on with some of the team's premium stuff, which is AI driven, but not co pilot. But just to be clear, I haven't had hands on but lots of my Microsoft friends have. And obviously, I've done a lot in chat GPT open AI like I've been using the tools outside, which essentially is the same type of experience. I'm really excited about it like Microsoft, people have really good feedback, we're yet to see how much the price is, which is always going to be interesting. But the potential of all the knowledge you have in in 365, your documents, your emails, your meetings, your meeting transcripts, being able to say to AI, oh, remind me what have I caught up with last on Ryan, it could span our chat, our email and our previous meetings. Like that is so so powerful, and just a massive productivity boost. I'm really excited for it.

Ryan Purvis 23:58:22
Yeah, I mean, as you as you said, I've used chatgpt a lot. I mean, it's part of my workflow now. You know, and but that's sort of jarring. And I think there's so many people that are not in sync with us that they are going to drown when it catches up with them.

Tom Arbuthnot 23:58:36
Yeah. And it's difficult with AI as well, because it's flawed at the moment. Like it's not perfect, it's 80% of the way there and understand like, like non technical people understanding that might be tough, because it's like, well, either it works or it doesn't it's like well, no, it kind of helps you. But you can't trust it. 100% you need to vet it, it's like well, like it like that, that kind of grey area of the whole hallucinations and imperfectness of it. I think like your typical user is going to struggle with that. And I think we'll see a lot of unfortunately, people generating stuff out of you know, generative AI, not even reviewing it and just posting it. You know, I take my word doc making into a PowerPoint, I'll throw it to my boss, and you haven't checked. Does it actually make sense?

Ryan Purvis 23:59:17
Yeah. Well, I've seen that we had it, we had a guy working for us. He's, and he's, I mean, he sent it to us. And he's like, alright, this talk last night, and he took me on that. I was like, You do realise that some of these sentences are, like, don't even sensible. Yeah, they write the sentence makes sense on its own, but like the paragraph doesn't make any sense. But did you generate it? Yeah. Okay, you know,

Tom Arbuthnot 23:59:40
what, where I have found a super valuable, we take our briefing videos, and we run it through first summary. So we take a 15 minute video transcript. And that is amazing, because we used to do that manually. And it took ages. And it will give you a bulleted summary, again, you have to tune it, you have to watch the video, but it'll get you a I'd say 80% 70% of the way there, which is a huge, huge time saving.

Ryan Purvis 00:00:01
Yeah, yeah, if I can go back to our sort of automation challenges. So I was trying to do something with auto where it'll transcribe it was transcribed and the meeting, but then I want to run that to check GDP to give me the summary and the bullet point actions because it doesn't do that pipe very well

Tom Arbuthnot 00:00:16
know that there's quite, there's quite a few tools coming in in that space. So AudioCodes have one for teams like there's one called crisp AI, which was a noise suppression app that now does AI summaries Gong in the sales space. And they're all trying to take meeting content and provide you minutes actions. So I think that will get better over time. But it's going to become the norm like at some point in the future, which should be like Well, obviously meetings generate action just in summary is like you don't do that the tech does that for you.

Ryan Purvis 00:00:46
Yeah. Now my dream I mentioned we have other episodes is that while you have in the meeting and you're discussing, like, we need to research this thing that AI has picked up on that it's already done the research and before you finish the metting, it's already told you what the what the

Tom Arbuthnot 00:00:59
weather? Yeah, yeah. Well, the thing that thing that I'm really excited about is it could be like, a really good PM. So I get like, people always commit to things in meetings. And then you have the next meeting. And they don't, again, like being able to get a list of actions, put it into a planning app. Like follow up before the next meeting to be like you committed to this last meeting, the next meeting is in three days like that, that thing that good PMS do really, really well could a lot of that could be assisted by AI. And we're not going to get that thing where we just have loads and loads of meeting saying the same thing, because they'll just be like, well, the same action list as last time.

Ryan Purvis 00:01:32
Yeah. And the I mean, is that thing already inside of exchange offers musics five, where if you've certainly emails where you might have read something, or commented on something, and they'll come back and say, Did you remember to reply to this? Yes. Or something like that. So

Tom Arbuthnot 00:01:49
there, that's the technology starting to come in? Yeah, you'll see it on LinkedIn as well. It'll be like you said this three days ago. Do you need to reply that kind of thing? Like that's very early use cases of that AI trying to pick out actions and commitments and timings?

Ryan Purvis 00:02:05
Yeah. Yeah. So I think it's exciting. If you had to sort of, I mean, do you ever feel that some of the stuff will be included in the market pricing? Or do you think they're going to? Oh, yeah,

Tom Arbuthnot 00:02:17
I mean, I don't have any inside knowledge, nothing public, but it will be extra for sure. Like the way they're positioning it, it's not going to be bundled. And then and there's a couple of clear reasons for that. One is this is very computationally expensive. So Microsoft need to pay for that compute. They've just dropped, you know, 10 billion into open AI, that's not charity work for them. They're looking to make a profit at some point. But also, like they're, they've bundled so much value into office 365. Over the time, they're now looking to add revenue. So you look at things like teams premium, it's an extra SKU that some of the things, phone teams rooms, they're upsetting things there, they will want to upsell this, and I'm paying for a few different AI tools at the moment, like, there's obvious demonstratable value there. If you're saving somebody, even half an hour, a month, an hour a month, you can start to justify a number for what that saves you. If you start to think about how AI could span departments and give you insights, like you know, things like fever sales or contact centre, you're talking about really big value there. So yeah, there'll be there'll be a cost to this. I'm not sure what the model will be yet but we'll probably find out is Microsoft end of FY, this coming up and then July is the new FY so we'll start to see the targets and the objectives in July.

Ryan Purvis 00:03:33
Yeah. So what are they have seen obviously, there's a lot of GPUs shortages. Yes.

Tom Arbuthnot 00:03:41
That's why Nvidia is now worth a trillion dollars.

Ryan Purvis 00:03:45
But I wouldn't, I wouldn't I wouldn't be surprised if Microsoft comes out with an acquisition in that in that space.

Tom Arbuthnot 00:03:51
Yeah, potentially. Yeah. But I think it'd be very a it'll be a very hotly contested space because you you look at Nvidia running the show there. AMD have done a lot of work there Intel have been caught pretty flat footed. But like, I mean, you can't buy an Nvidia or an AMD necessarily without a really big chequebook. And do you want to? The other thing that I think would come up really fast is the anti competitiveness, it's all very clear AI is going to be the future of a lot of things. Do do they want a chipset vendor living with a Google or an Amazon or Microsoft? I think that will be a challenge to get through the competition Commission's To be honest,

Ryan Purvis 00:04:29
yeah. When I say acquisition, I was thinking of a big player, but maybe a small player there was like what Apple did when they started the chip stuff again, they bought a couple of small companies. Oh, yeah.

Tom Arbuthnot 00:04:38
Yeah, maybe there'll be some orientation like and obviously Microsoft have a bit of a hardware history with X Xbox and with some of their devices that they're not it's total strangers to hardware. Apple's the interesting one I wonder what there hasn't been a lot from them yet, but they're doing some pretty amazing things in the processor space with their M chips like at some point there'll be some cool stuff coming from them in this space as well.

Ryan Purvis 00:05:01
Maybe I look at their headset, I only

Tom Arbuthnot 00:05:04
the press release, obviously. But yeah, interesting. I mean very high price point which immediately knishes it out to the to the very high end. I love the fact that they've new like got the whole you don't need glasses, you put lenses in it type thing, but I mean, Gen gen one or anything like this will be the get one to market. So we'll see see how it how it lands. But I think it's interesting that that kind of technology has been around. You know how much of that Highlands Google had lost. And what Apple are very good at doing is not being first but then doing a really crafted experience and I suspect that will be them again, there'll be like it will be slick If it will have consumer application, so it will definitely move the market forward. But I think that whole meta augmented reality thing has taken a backseat to AI for most people at the moment.

Ryan Purvis 00:05:56
Yeah, yeah. I mean, I think they are we our use cases are still there. And I think there'll be solved, but knowing I think they asked me is far more interesting. And I think it hits more people. Yeah. You know, if you think I'd be a PowerPoint monkey, or the Excel spreadsheet, you had to do you know, that kind of stuff that you can just, oh, yeah,

Tom Arbuthnot 00:06:14
that's that's most like it's it's most knowledge workers get a benefit. They were there's only a subset of people that want to wear a headset and do that thing.

Ryan Purvis 00:06:23
Yeah, exactly. And if I look at, you know, just from sort of the the use case of we're having the meeting, we're discussing something, or they are already doing it. The other piece of there is what you mentioned around the sort of business process improvement pieces. So you're writing a document, you're writing the gist of the document that gets turned into a full document, or some or not even forget that to summarise to just the same way that everyone can understand the multi language, and that kind of stuff, that that's the huge value piece. And then that changes your conversation with using translation services we're using using legal services, their attacks and all that kind of stuff,

Tom Arbuthnot 00:06:58
the space that I think is gonna get disrupted, the quickest is contact centre, or kind of customer contact. Because it's really interesting. It's such an obvious use case of AI, which is, you as a customer are trying to get a problem solved. Like 5060 70% of time, it's the same type of question, yeah, open hours bank balance, whatever it is, it's perfect for AI to solve those problems. And it's human wise, it's expensive. And that the bar of quality there is challenging in terms of user experience, like, it's not difficult for AI to be a better experience and your typical human agent. So I think that'll get disrupted really heavily. But also contacts into vendors typically charge 150, you know, $120, a seat? How does the AI bring a lot of the cost out of that? Did do they try and charge for a virtual agent, they charge for resolution? It's really interesting. But I think if you look at industries that are getting disrupted quickly, that will be one of them, someone will come up with an AI that can answer 50% of the questions that sits in front of the people, and it starts to answer them, be that chat be that voice. And that's an immediate savings to the organisation and probably a customer site improvement as well.

Ryan Purvis 00:08:07
Yeah, and I don't think you necessarily lose the person in that process. But if they're gonna answer 50% of the questions, and do it via an easy interface, wherever that is, but then when you don't get your answers to those questions, and you talk to an agent, that agents got more time,

Tom Arbuthnot 00:08:23
yeah, the agent will become a high school job, not a lower school job, it will be like these people are the problem solvers. Like they're the people that can help customers when the technology can't help them. So you'll need less of those people. But there'll be absolutely higher skilled. And also part of that team, or that role will be putting content back into the engine to solve it the next time. They're like, Oh, this is how I solved that problem. So the AI gets better and better. So yeah, I human humans in that scenario, don't go away. But they do get radically reduced numbers wise, and that's been happening already with chat bots, things like that. But I think the fact that this can be done with voice will take it to the next level.

Ryan Purvis 00:09:01
Yeah, and I mean, a lot of those those API's do a very good voice systemization that

Yeah, I mean, we were doing some some demo stuff about a year ago. And none of us wanted to do the voiceover for the demo. And one of the guys has wrote a script, and he put it through an API and it generated a nice, quiet, there was an American female voice or an Australian, wasn't Australian. It was I think it was an American female. Yeah. And, and it was, it was great. And a bit of a joke for those take the Star Trek people, but it used contractions, which, you know, daily, daily couldn't do during next generation. But I was it was, it was hugely impressive, because, you know, there was no stuttering it was nicely spaced, so there was no in rush, that kind of stuff is, was

Tom Arbuthnot 00:09:48
there's a tool, we use it empowering cloud for editing, it's called descript. And it takes a video, and it transcribes it to text and you can edit the text, and that edits the video. So you can take out filler words, like it's amazing. But it also has a text to speech that you train on your voice. And you can type and it will do in your voice. So and if you put time into training, that it's really good, like, the things like you know, you make use, I should do this, rather than shouldn't do this, you can literally just type the replacement and it will fill it. It's pretty impressive.

Ryan Purvis 00:10:18
Wow, I didn't know that. But that does line up with something else that Apple was doing where you could record your voice through Apple to preserve it.

Tom Arbuthnot 00:10:26
Oh, interesting. I didn't see that. Yeah,

Ryan Purvis 00:10:28
I'm trying to think I heard that. But basically, you have like 15 very odd bits of things you've got to say and then you can use that. I suppose you can use the voice to speak to somebody else. I guess the and obviously as you get older your voice changes so you could always there'll be no really recorded from prosecute ban.

Tom Arbuthnot 00:10:46
Yeah and in nuanced do that for security. So you can do bank passwords and a factor is recognising your voice. So like there's there's lots of use cases that voice signature.

Ryan Purvis 00:10:57
Yeah, we were doing it for password resets. And then I think one of the medical aid companies think it's discovery in South Africa, you can short circuit the whole IVR queue, if it's got your voice record in place. And that's a huge value, because going through the stupid IVR systems is

Tom Arbuthnot 00:11:15
I'm curious now with that, that voice recognition how good the AI is as to whether the that will become a problem now, because you can technology becoming readily accessible to dupe somebody's voice, obviously, training those engines, you have to have a large sample set. So for your typical person, not a problem, but maybe you either have hours of content lying out there and might be a problem.

Ryan Purvis 00:11:36
Well, yeah, I mean, I don't know if so there was that one influencer? She, she used AI to, to do this. And she was charging something like $77 value per minute, and she made like, 70 grand? And of course, because because it's an AI, she wasn't doing the interaction with the people. Yeah. My expectation is that a week down a pretty sorted route. But, you know, yeah,

Tom Arbuthnot 00:12:03
I've got a theory on this as well, which is that like, content generation is going to get easier. So you're gonna get all these crappy blogs of like marketing, people taking top 10 teams tips, bang on the website, job done. And I think people are gonna come back to personalities that they follow rather than content they follow. So it won't be, you know, you read X Y, Zed magazine, or X Y, Zed blogs, you read Ryan's posts, because you know him and you know, his opinion. So I think there'll be a resurgence of people looking for individuals for information rather than generically googling it, because all the top hits are going to be AIG and perfect articles, but no opinion, no, no substance, just heavily crafted wording. Yeah,

Ryan Purvis 00:12:46
that's, that is the vicious circle, because you if you got a nice chat now to go and give you an SEO generated article, it's going to be giving you basically it's going to do that for a whole lot of people. So everyone's gonna have articles at the top. So it's really gonna be probably the time that you publish, it gets you to the top versus the, the SEO conference level, let's call it soccer really, though, like I could see YouTube and live streams and Twitch and all those things becoming more more watched. Because people want the real they want

Tom Arbuthnot 00:13:18
they want to hear from that person. They want to hear something real, not something generated.

Ryan Purvis 00:13:22
Yeah, yeah, exactly. Exactly. And I did wonder about all the all the hidden tags within the generated stuff. So the voice or image when it's generated?

Tom Arbuthnot 00:13:36
Yeah, there have been there's been governmental talk about do we need to force the technology companies to do that? I don't know if there are or not, but there are there are there are recognition, things that you can run tech through backwards to say, Does it feel generated or not? But they seem to be, you know, hit and miss. But even if the government enforced that the sound is out there now and widely available, the people that are bad actors will will use variants of stuff that doesn't do it. So I feel like that ship has sailed. But on the other hand, as it gets you right now you can recognise it, you know, someone doesn't need a video generated Talking Heads, it's the the head keeps like moving like this, you know, in a constant circles, make it feel real or whatever. But that will all go away, and it will become much more engaging. It does beg the question like, do we need to have to flag that this is AI generated?

Ryan Purvis 00:14:22
Yeah, and I think people don't realise this. But if you're using generative AI, all that content that you've got is not is not yours. It's copyrighted already, from the source where it was generated from, or something that effect or the engine that generated it gets the doesn't own it. But but it's, it's not you can't really use it can Yeah, you shouldn't be able to think, yeah, I think you've always even though you're generally still gonna take the generic content and still make it your own. Or it's done as giving you volume.

Tom Arbuthnot 00:14:50
Yeah, I mean, what's the problem is what's the definition? Because like, if you use Grammarly to correct your work. I mean, that generates suggested sentences. So does that partially generated it like, Oh, is that where's the line on? What when does it count as AI generated versus AI assisted? versus no AI? Because actually, all your all your word processes spell check now that's all AI. So is that AI and AI? It's a it's a hard line to define.

Ryan Purvis 00:15:20
It's a good point. And I think we're gonna see some court cases that are probably set those boundaries and I think unfortunately, like anything, and you mentioned the speed of things progressing, the laws and the policies are knowing Yeah, yeah. Moving at the same speed. So

Tom Arbuthnot 00:15:34
that's a big conversation I'm having with enterprises at the moment actually is the simultaneous the excited about AI and really fearful about what What if all data gets leaked into it? What if someone like what if a lawyer generates a contract partially from Ai? are they accountable for it whose account? Like, there's a whole bunch of alongside the excitement worries about governance and compliance and risk and and just a general worry that people are gonna over lean on it and not be as creative in their roles, like you're paying them to come up with a solution if they're just using AI? Who's that good for really?

Ryan Purvis 00:16:11
But it's funny, you mentioned that that lawyer thinks so there was actually a case in the states where a attorney was I think he's been disbarred, because he used chat gpt to generate all of his case, submission, whatever. I don't know the legal terms. Yeah. And chat gpt to be referenced a few cases that didn't actually exist. Yeah, position picked up on this and said, Well, you need to you know, what happened, Janice log in, I use chat gpt, in essence, and while you can't, you know, that's basically the I would say it's fraud. But But beyond a lot of job.

Tom Arbuthnot 00:16:42
Yeah, yeah. It's a problem, isn't it? Exactly. It's, yeah, well, I mean, the genies out of the bottle, now we are going to be using this stuff in various ways we'll find new norms or so it's just a case of, it's a really exciting time to be alive in the sense of, you know, I think the again, we're in the we're in the middle of the hype cycle. But this feels materially impactful to most people. So it'll be interesting to see where we are in a few years.

Ryan Purvis 00:17:08
Yeah, I mean, unlike with AR and VR, where that was a hype cycle, there was always going to burn out. I don't think Yeah, I think there's too much that could happen here.

Tom Arbuthnot 00:17:21
The other thing that is that makes me fairly bullish on how impactful it's going to be if you look at some of the conversations around how faster compute skills or the compute capabilities moving so to generate check GPT 3.5, that data set on the original computer they used? It was I think, was it 40 million or something like that the cost to generate that model is a fortune. And in two years that has dropped, like, you know, a few few 100k, and it'll drop down, it'll drop down as the computer gets better. So that will make the model creation more commercially accessible to everybody. And you can see this already, it's very real. I think Bloomberg is taking all their data, and they're building a model around their dataset. So I don't think you'll have one AI to rule them all. Or maybe who knows, but but certainly what you'll have, as you will have the the AI for all the business questions you've got based on ft and Bloomberg and blah, blah, and they'll own that piece. And you'll have the AI for car parts. And there'll be some tool that somebody builds that has every car part model and how you order them and that kind of stuff. So rather than seeing this one giant AI that wins the race, I think we'll see lots of specialised models, which will be exciting.

Ryan Purvis 00:18:31
Yeah, I mean, that one wants to reward is very much the Skynet, or the science fiction view. There's a lot of books that have been really recently funny enough. And they're written long before this hype cycle. So it's been interesting to read those and be like, hey, and he's actually pretty, pretty accurate. Yeah, prayers are what can be happening. And you know, one of those things was deep fakes. And the other one is, is a future where people have gotten so used to AI is doing everything for them, they actually don't know how to do anything. Yeah. And, you know, that's reality. I think a lot of people you can look at, like writing code, for example, if you look at a developer who started writing in C++, or C, or some or something along those lines, you know, now you get to a dotnet. C sharp with so much stuff has been abstracted away.

Tom Arbuthnot 00:19:18
Yes, you have no idea what the hardware is doing. You're just doing the code that the code says, Yeah, it's interesting. Like, it's definitely a race, I think we have to, certainly the next generations, we have to teach them how to use it as tool, not rely on it. Like, like, I see some schools that are trying to ban it. And I see some schools that are trying to say, Okay, this is like Google, like we teach our students how to generate a base paper, and then they edit it, correct it like that's their task, which I think, you know, with all technologies, we know that banning rarely works. I think the model is help people understand where it's at all and how they use it, so they don't over rely on it.

Ryan Purvis 00:19:56
Yeah, 100% I mean, I think that the critical thinking skills, the problem solving skills, the ability to interpret something and see if it's right or wrong is the value. And I hope we don't go down a route away because now we got this ability to generate content. There now you get double the amount of assignments will tend to

Tom Arbuthnot 00:20:15
Yeah, because it should take you half the time.

Ryan Purvis 00:20:19
You know, but I think that time will tell but I think education generally has to change anyway they are still doing things. That's old factory mindset.

Tom Arbuthnot 00:20:30
I quite like Microsoft to give them some credit. This copilot branding is a really smart branding which is not the AI does it for you. It's the AI assist you their landing well that message and it really helps them from a Training point of view and a tactical point of view that you know, this is you are in control of it helping you you are ultimately accountable for it, you are using it as a tool. That branding is really spot on for how to position the tool

Ryan Purvis 00:20:58
as genius and I like to use the word landing to stop use the copilot, which was, which was awesome. Oh, yeah. Yeah, I think you're right. I was walking. Think about it yesterday, that is the right way to do because often, we sit even before even before someone buys a tool, and they expect the tool to solve all their problems automatically, magically. How many times? Have you heard that I bought this tool that it doesn't do? X, Y and Zed? Yeah. Yeah. And I think that's, you know, the reality we should be looking for is that you've got a tool that helps you get x step ahead with whatever value it is, and you still have to do

Tom Arbuthnot 00:21:36
what Yes, yes. saving you some of the time, not all of them.

Ryan Purvis 00:21:41
We met some guys were in Egypt, this last week, we were in a radiologist. And I was just asking them, you know, what are your thoughts on AI processing imagery? They said it was pretty good. You know, for chest X ray. It's like really accurate, and it picks up a lot of stuff. But it's still missing some stuff. Yep. And but if it gets to 80% of them, and if you do 20? You know, I'll take it because, you know, yeah,

Tom Arbuthnot 00:22:04
or if or if 80% more people can get access to a charter that being found, because there's only so many ratios, looking at charts. And you can scale that out is the other way to think about it. But 100%.

Ryan Purvis 00:22:14
And he was saying, like, from a from a job security point of view, he doesn't feel less secure. He feels more secure. Because he's adopted, he's using it and he knows, already knows what he has to do to compensate for. Yeah. So it goes back to the copilot thing. I think it's you know, it's a partnership, much as we are.

Tom Arbuthnot 00:22:34
Yeah. So anyway, tell me what's going on in your world? What's What are you excited about at the moment?

Ryan Purvis 00:22:38
Well, Valuu is very exciting for me. So obviously, we have rebranded. Well, Digital Workspace will stay, because that installed great platform, and we get a lot of good guests, and we talk about stuff that I still care about a lot. So it's Heather and Catherine, who's who's joined us recently, but values the thing, and it's very much along the lines of what what you're doing. So we have what the community thinks will be a new thing that that's recently become a thing, and that's Valuu execs. And that's a lot of CTOs CIOs that are looking for opportunities to help on a on a part time basis as a fractional, you know, businesses and, and it could be startup, it could be, you know, businesses looking to pivot. Or it could be the enterprise, it was actually an interesting thing. Just going on this one for a little bit. I was someone said to me what fractionals are only for startups and for small businesses looking to grow. And I said, well, not really, because if you think about it, now, it was to your point earlier about how fast things move. If you're in an organisation, and you need advice on how to bring machine learning into your ecosystem, you could go and find a big consulting firm or a consulting firm to come and help you. Or you could tap the shoulder of CTO CIO equivalent or C level person that you know, from your network, you could come in and help you coach and guide and reorganise before you bring those consultants in. So there's still I think there's an option there too.

Tom Arbuthnot 00:23:56
Yeah, definitely. I think I think you're right. I mean, you look at the proper consulting, I guess, as opposed to selling big consulting projects, like the idea to come in and provide some expertise. On a part time basis. There's a winning both directions there, because they're getting visibility of another organisation another use case? Absolutely, I think there'll be more and more of that. And it's really interesting to do it, cross customer. So we do service owner roundtables, and we bring together different verticals. And that's always interesting, because you're bringing someone in healthcare with some of the legal with someone in pharma and being like, Oh, we solved it this way, we solved it that way. So getting beyond your immediate vertical was really interesting as well. And that's a good place for fractional.

Ryan Purvis 00:24:35
Yeah, 100%. And a lot of the people that I'm talking to is at that stage in their careers where they've built up multi industry, or multi vertical knowledge. And they've seen a lot of this transition that we were experiencing, you know, from from a big organisation or small organisation point of view, and they just want to share that knowledge. I just want to help you Oh, look, I mean, there's a commercial angle to it, we're on nausea for free, per se, they want us to pay the bills, but, you know, the opportunities are there to help somebody else. And that's why we sort of positioned as a stepping on the shoulder of giants because you don't have to go and learn that stuff. Because you can talk to someone who's got the expertise, you're not going to pay them the full salary or the full wage, you pay them fraction in which Yeah,

Tom Arbuthnot 00:25:17
yeah, there's also a whole mentorship part that which is beyond just the knowledge, which is they've done it before in a different place, and that the people working with them get value from that relationship as well as just the knowledge and the insight.

Ryan Purvis 00:25:28
Yeah, and we're just from a value point of view, we're looking just to connect we're not we're not really looking at the getting involved in the commercial agreements or anything like that. And some guys are willing to do if equities like a startup, they'll do equity or consider getting paid and all that kind of stuff. So that's, that's a new thing that that's going to be I'm pushing out that at the moment. And then the core things that I've been focusing on, obviously, the advisor work that I do, helping customers find value in the digital transformation projects. They're either when they're still planning and thinking about what to do, or they've bought something or they've run a project. And they're like, we're not seeing any value out of this, that I come in and help them to, to figure that out. And so they're still there, that consultative thing that I do, which

Tom Arbuthnot 00:26:07
we saw before, I mean, that's a never ending problem, isn't it? People will spend, you know, 1000s, or millions on projects, and they're like, it hasn't quite turned out the way we expected. And often, vendors are very excited to tell you the latest thing, but where was what what was the original use case? What was the ROI driving the value out of projects? And as the people get a little bit more cost conscious, economics wise, there's more focus on that. And it is not in the interest explorer space, we've already bought this stack, that we're not using half of it, how do we drive the value of what we're already subscribed to already paying for?

Ryan Purvis 00:26:39
Yeah, I know, we march that's a great one for that. Because there's so much stuff in there. So many different products. So many slants on a project, like which one is the right one to use. I don't know. Is it Logic Apps? Is it power automate? I don't know. You know, it's some of that stuff. But then this is where the SAS product comes in, which you've built as well. And that's going to be that's we're launching that now, if anyone's interested. So this is taking the the methodology that we use, and allowing, but given the ability for people to use it themselves. So how do you prioritise the work? How do you pick up on the scenarios you want to address? How do you calculate the value? How do you so from the beginning, like, like building your business case, to the end, where you've now run the project or through the project, you're comparing the whole time on what your value driver was? And how you tracking? So that's, that's the part that I'm most most excited about. But that was the dream from

Tom Arbuthnot 00:27:25
taking it from a from a consulting based scenario to a framework and the system. That's really exciting.

Ryan Purvis 00:27:30
Yeah. And then we've got that little bit of training that goes with that about what's the methodology, one of the things that we think are useful skills for people to have. So we've got about nine little courses that are going to come up now. So it's this whole little ecosystem of value. And then I was looking

Tom Arbuthnot 00:27:45
at the Who's the person that's for it, like, what's their role? What's their, what's their goal?

Ryan Purvis 00:27:52
So I see it as the consultant, person, or be all company that is there has got a customer, and they need to help them with an initiative to find value. So they're planning, they're tracking and then realising the value. And so that could be a consultant could be a one person, business, or it could be a few people, whatever it is, and then you get into big organisations. I also see it for the system integrator, the managed service provider, someone who's doing customer success, someone is doing pre sales, where they are dealing with the climate factor that email literally just opened up on my thing with we saw a report where the guy's got to deal with a six figure deal. He's asking me like, how do we do a value? Pitch for this? Yeah. And it's, you know, but that's

Tom Arbuthnot 00:28:41
really, really smart. Because also your, the one of the biggest problems at the moment is people get sold things, they don't execute on them properly. And 12 months later, because everything's SAAS, right, that they're like, Oh, well, we're not, we're not going to renew this, because we're not seeing the value, we thought we were going to switch to the next or the next s. So helping your customer on that journey from presale to customer success and having a framework or is really smart.

Ryan Purvis 00:29:04
Yeah, and that's exactly where it. So that's the originally I was trying to sell it to the customers, the actual big enterprises, and one of the guys that I spoke to know you've been silly, because that lead sales cycle is gonna be so long. So that's why we're now focusing on sort of consultant. Persons, I mean, a customer can come by it, if they want to. And, and that's the this is where the, the ecosystem works quite nicely, because we will have a partner model and we'll drive around with your partner. But you could through the system request a fractional sort of the value exec, or you can request the advisory person and for the for the person who's using the platform. That's a that's a free transaction. Just tell us what the problem is. We'll come and help you and then the commercial thing, you know, we'll figure it out as we go. Because we just want people, you know, to get good to get the value they're looking for. Yeah. So yeah, very exciting. So that's really cool. Yeah. Thanks for asking.

Tom Arbuthnot 00:29:58
Yeah. So you've had the same fun I have of all the joys of platform building and developers and all that good stuff.

Ryan Purvis 00:30:05
Yeah. Look, we've got some good guys. They've done some great work. And we've been using our products, a low code, no code product called bubble. Oh, yeah, heard of that. Which, which we built in really, really quickly. I mean, there's some there are some nuances and that but that's always going to be a trade off when you go low code, no code. But, you know, I can't you know, if you'd asked me a year ago, what would this thing look like? And I knew that I had to build it from scratch or through a platform the normal way. You know, two years later, we'd probably be still building it. You know, now here we are. It's almost ready. But it's in pilot with a few customers already settled. So it's been used. So when I say it's almost really, it's it's literally the like the last things like how do you log a support case? How do you add a feature request? It? We will finish it.

Tom Arbuthnot 00:30:53
I have a look at intercom intercom is amazing. That's what we use for our support ticketing and chat. It's so awesome. I couldn't can't speak highly, more highly of that product is great.

Ryan Purvis 00:31:03
Yeah, he's at a high dose. Very impressed with that. We're using HubSpot at the moment. But yeah, intercom is always on my list. It's one of my go to

Tom Arbuthnot 00:31:10
Yeah, we did some stuff. We have spoiled the pricing scales really quickly on that if you want the bells and whistles that that was another thing where we were like, Oh, this is good. And then we're like, oh, we want API access. Oh, it's a trillion dollars. Great. Thanks.

Speaker 2 00:31:24
Yeah. Now we've had that challenge. So we're giving it we're giving HubSpot a year. I think I bought it last year, December. And we've looked at a few others. And I've been impressed with with the likes of like Zapier, for example. They're just building out some of the most interesting things. So for example, I used to use notion as my, like, if I was going to do something, he would go to notion and suddenly and then I would kick a workloads from notion. They've built that in that table thing that exists in Zapier. So I wouldn't be surprised in X number of months, because I think it's

Tom Arbuthnot 00:31:57
Zapier becomes the product as well as the connector.

Ryan Purvis 00:32:01
Yes, I think it takes over a certain space and product. Yeah,

Tom Arbuthnot 00:32:04
notion is interesting. I use notion a lot like it's a bit of a challenge, because it's a blank canvas. So you can spend as much time building your system as executing your system, which is one of the challenges we went through, we use Monday for a while now we're using clickup. Like, none of them are perfect, but clickup has a really good mobile experience. So we're quite impressed with that.

Ryan Purvis 00:32:23
Yeah, so and this is the new thing. So when we built the value workspace, which is the SAAS product, we didn't want to get into the whole project management side of things, but you kind of have to because you're doing governance. So we both have our version of trying to keep it as simple as possible. And one of the things that we don't have, which I think we'll have to do is the mobile experience at some point. But I don't again, I would rather integrate to to something else that does the project management piece. Yes. And all that.

Tom Arbuthnot 00:32:48
Yeah, the API coverage on those tools is really good. So in theory, you could pump your actions, tasks, follow ups out to that third party system planner, Monday, click out whatever, that might provide more value to the end customer because they're already living in that world rather than having to check your tool.

Ryan Purvis 00:33:05
Yeah, yeah, exactly. And the stuff we're focusing on is more on the bit on the Generate the the value, realisation and predictive piece of that, which is a whole different discussion. But it's exciting, we're seeing, I'm seeing a lot of it on LinkedIn, funnily enough, people looking for these sorts of solutions. And again, a lot of messages there. And I just think it's a case of of finding what the user is looking for it is finding the different scenarios that help us to to close out problems, because that's really the trick is what are all those scenarios? And how do you solve them?

Tom Arbuthnot 00:33:37
Yeah, it's actually the your approach is widely applicable. So it's finding the right use cases that are emotionally make sense for the customer, for the supplier like that you can drive a result. And the other thing is driving a result in a timeframe, because everybody wants things yesterday. So like, being able to show customers, we did this thing with Salesforce or Trello, or whatever it is that the organisation has invested. Now they're seeing value out of it. And this was the framework, I can apply that to your project is slightly different. But the fundamentals are the same. What did you want it for? How do we drive the engagement, the activity, whatever it may be? It's, it's really exciting space?

Ryan Purvis 00:34:14
Yeah. As you say, it gets me up early in the morning and keeps me up late at night. So great. So with the new business, I mean, do you want people to go to Empowering cloud, or you wanted to go somewhere else just to have a look? Yeah, if

Tom Arbuthnot 00:34:26
they're, if they're interested in martial teams, or modern work, if they had to empowering dot cloud, they can log in for free with LinkedIn, or an email address. And there's a whole bunch of free content out there from MVPs NCAM people in the community. So our whole goal is to keep you up to date on that Microsoft modern work experience. questions or feedback hit me there on the forums or the intercom. And yeah, if it's if it's Microsoft partners, or vendors or IBM's, who want to train their teams and kind of drive that through then there's a conversation there too. But I'd love people to check it out and always looking for feedback.

Ryan Purvis 00:34:58
super well. Thanks again for coming on the podcast was great to catch up.

Tom Arbuthnot 00:35:01
Yeah, really nice to catch up cheers Ryan.

Ryan Purvis 00:35:05
Thank you for listening to today's episode. Hey, the big news app producer and editor. Thank you, Heather. For your hard work on this episode. Please subscribe to the series and ratings on iTunes or the Google Play Store. Follow us on Twitter at the DW W podcast. The show notes and transcripts will be available on the website https://www.digitalworkspace.works/. Please also visit our website https://www.digitalworkspace.works/ and subscribe to our newsletter. And lastly, if you found this episode useful, please share with your friends or colleagues.