Succeeding in an AI World: Vital Training, Bias Awareness, and More with Waseem Ali, Part 1

00:00 - Waseem (Guest)
What individuals need to do is learn to upskill in these technologies, because actually, their role might broaden as a result of bringing these new technologies into play, which is why we're seeing so much demand for people around AI, literacy and so on. Right, is there certain roles that may go as a result of AI? Yes, but that's existed for years.

00:21 - Mike (Host)
My name's Mike Lander and you're listening to Higgle, the B2B Sales Club podcast, where we bring you actionable insights about sales, RFPs, negotiations and difficult procurement discussions from sales leaders, brand leaders and procurement leaders. Please subscribe to get updates when new episodes are released. Waseem, thank you ever so much for joining me on Higgle, the B2B Sales Club podcast.

00:47 - Waseem (Guest)
Thank you for having me, Mike. It's been a while in the making, but really excited to be here today.

00:52 - Mike (Host)
It has. I think we've both been busy and we've bounced it a couple of times, but thankfully we're now in the same place. Well, different places, but at the same time. Some with newly decorated rooms In a newly decorated room exactly. Some with newly decorated rooms In a newly decorated room exactly, with the nice artwork at the back. It's very good. And the nice colored walls. Exactly, just talk about the artwork. So who are you, what do you do and what's your favorite song, and why?

01:18 - Waseem (Guest)
Let me start with an easy one as to who am I and what I do, and hopefully in the background. Think about a song as well. So who am I? My name is Waseem Ali. What do I do? I'm a data person and have been for many years.

01:38 - Mike (Host)
I think it's fair to say it's the only thing I'm remotely good at.

01:41 - Waseem (Guest)
I think I'm sure that is not true, waseem. Well, if start start, go go years and years back. Uh, the only thing that I was good at school was maths and statistics, right it, for some reason I understood it and it just clicked and it worked. Everything else I'm not very good at and that's kind of translated into the degree that I did and it translated into the career that I I've taken in in data because essentially I'm taking, I'm doing what was a statistician used to do, and taking data and do it now.

02:07
It's not a big, big field, right. So I have my career started off in healthcare and in consulting. I've done a little bit of business in between. I went into kind of a leadership role at a relatively young age, managing teams. I was very fortunate to get the opportunity to do that. I dared to start my own business, which I was fortunate enough to sell, in data strategy, consulting and training. I was prior to the role that I'm currently in, I was the chief data officer at Lloyds of London where I built and stood up a team of 150 to 200 odd people managing you were the chief data officer at Lloyds of London.

02:42
Yes correct.

02:42
Yeah, okay, that's not a small role, is it? No, it was a big role. It was a very exciting In a highly regulated environment Correct, yeah, and we had all the challenges from GDPR at the time to implementing a new digital and data strategy, to building out the team, to rolling out data literacy for the market and the corporation and stuff as well. So that was a really exciting time for me and the corporation and stuff as well. So that was a really exciting time for me. So I kind of went from startup to big corporate, regular corporate to back to startup with my current business. So I had the privilege of being the CEO of a company called Rockborn. We are a data and AI consultancy and training company with a slight twist. So we pride ourselves on training early careers talent in our academy and then we deploy them as consultants to client sites. We provide B2B training services for end clients as well, around everything to do with data, AI and the human skills needed to be a data and AI person at the same time, which is really interesting.

03:36 - Mike (Host)
We're going to come back to that definitely.

03:38 - Waseem (Guest)
And consulting what you would expect an organization to do, so we can come in and we can do a delivery for you and we can deploy a squad, and so on and so forth. So I've been been ceo for a couple of years there, gone through its own transformation as a business and and getting it to the point that it is today, but where we're starting to build quite a nice reputation in terms of what rockborn does and how we are kind of a very high quality service in the market that we operate in as well. So that's's a little bit about me. I don't know if you need hobbies and stuff, but I love playing tennis.

04:07 - Mike (Host)
No, that's fine, there we go. You play tennis, okay, that's good. What about the song? So back to the song.

04:12 - Waseem (Guest)
Well, I was hoping you didn't remember, okay.

04:15 - Mike (Host)
Oh no, I'm afraid not.

04:19 - Waseem (Guest)
No, no, no, no. Listen to the most is is a song called fix you by Coldplay. Oh yes, I've got attachments to the song due to certain things that had happened in in my life at the time, and also I'm I'm a fixer. I'm annoyingly a fixer. If something goes wrong, if I can't fix it, I really struggle. So so that's that's. That's my own kind of demons coming out in a way. But, um, one song that I've probably listened to the most not recently, but definitely one that I've listened to the most- so the fixer thing is something we definitely have in common.

04:49 - Mike (Host)
So I'm an engineer, by training an electronics engineer, and I still like if something's broken I'll try and work out what it is and I'll try and work out how to fix it. And sometimes that drives me mad and the family mad, but I enjoy it, I still enjoy it. I still enjoy it now.

05:01 - Waseem (Guest)
So interestingly and this could be a whole podcast on its own what do you do when you can't fix something Because you can't control certain things, right, health, whatever it may be. So what do you do in a situation where you can't fix things?

05:15 - Mike (Host)
So that's very interesting. So I'll take an example where I think. So I'm a bit stuck currently. Yesterday's outage affected Ring, so Ringcom went down and my doorbell stopped working. But I didn't know Ringcom had gone down. Did a test Status said it was online but the door camera was working, but the bell wasn't working. So I reset it. Now that was a mistake, because it won't work now and ringcom's back up. So now I'm struggling with okay, how am I going to fix this thing? So the solution's going to be I suspect next stage is to call customer service at ringcom. If that doesn't work, ultimately I'll probably have to buy a replacement doorbell, but that will annoy me because I won't have been able to fix it.

06:10 - Waseem (Guest)
That's a great recent example, right.

06:14 - Mike (Host)
So let's get into the questions before everyone starts to drift off, thinking we're going to talk about doorbells because we're not. First question so you operate in this intersection of AI, training, consulting and resourcing. So talk to me about how market demand for these types of services have changed in the last year, two years. And I ask it because, as people that know me, I came from the resourcing background. I used to work in a very big recruitment process outsourcing company and a managed service provider, so they were deploying thousands of contractors at the time onto client site and I can see that that market's evolved from just deploying talented contractors into kind of what you're doing now. So just explain again what problem do you solve and why has that become a thing?

07:04 - Waseem (Guest)
Yeah, and I probably should have said this in the intro we're part of the Harnam group, so Harnam is probably one of the largest data and analytics and AI recruiters globally in this space and have been around for almost 20,. I think they're celebrating 20 years and we celebrated their 15,000th placement, so it's a very big in the specialized space of data, ai and analytics, right. So actually I've seen the impact on the recruitment market as well as my business as part of that, because I'm on the board of Arnhem Group. So actually it's been really interesting, right, because I've learned that in an upturn. So when the market is on the up, recruitment companies, talent companies, consulting companies are probably the first to feel the effect of it, but on the opposite side, on the downturn, they're the first to feel the effect of it as well.

07:52 - Mike (Host)
That's exactly right. It's the same as taxi drivers, it's the same bellwether. It's a very early warning signal Recruitment turns down, we know that we're in trouble.

08:02 - Waseem (Guest)
Absolutely right, so, actually. So what have I seen? It's been a difficult market. I think organizations have, like whatever phrase you want to use tighten their belts, not spending as much, going through restructures, redundancies, all of those kind of things. The stuff that's come up in the news has been is obviously for the big tech vendors letting go people, all of those kind of things. That's been an active part of the past kind of 24 odd months or so.

08:29
When we started seeing signs of that certainly from my business's perspective I always thought it was going to be kind of long and slow, if that makes sense, and maybe I was deep as grumpy at the time, but it turns out it has ended up being long and slow, and so what we have found is where organizations are not perhaps hiring or have tightened their belts in terms of permanent hires in their business. There's still a lot to do, because this has happened at the same time as generative AI has become a thing. Right, ai has been a thing for years and years and years. Right, but generative AI actually became a thing and it went through the hype cycle over the last kind of 24 months or so, right, so it's happened at the same time. So, whereas organizations are going through the economic strain of saying, oh, we need to tighten, we need to control costs, ai is happening at the same time and organizations are probably feeling like if they don't aren't a first mover or a fast follower, they'll fall behind, right. So it's been really interesting in that space, because what they've started doing is turning to businesses like ours for advice, market insight as to what other organizations are doing. They've turned to us to deliver projects, so actually going and deliver certain proof of concepts or alpha releases, beta releases, so on and so forth, or provide their organization and their employees with training around data and AI, and actually that side of the business has been very, very popular.

09:51
I'm obviously biased. I think we're very, very good at training and I declare my bias here. I think we're one of the best out there but actually what organizations have turned around and said is we need to upskill our people, and we need to upskill our people in these kinds of things. Can you help us design something? So what I've seen is there's been two things that I've seen in the last 24 months. One is we need to control costs, obviously, like every business does, but at the same time, we need to get onto this AI thing that organizations are doing, and step one is let's understand what this AI thing is. Is it good, is it bad, is it scary, is it risky, is it beneficial, what is it? And so there's been this journey that organizations have been on and continue to be on.

10:31 - Mike (Host)
So, waseem, help me and the listeners with. So when we say AI training, that obviously there's a big spectrum there, from engineering level AI training, data level kind of training, and then business people that need to understand what the opportunity is and how to unlock it in a safe way. Where do you think most of the training need is coming?

10:58 - Waseem (Guest)
There's two areas that we've seen a lot of demand and I think will continue to be a lot of demand. One is actually I ran this initiative with a colleague of mine at Lloyds in London, where we actually said we were talking about data at the time, but it's very relevant to answering your question we wanted everyone to be a data person. I wanted to move away from the. There is one central team that knows everything to know about data. Only come to them to hold up a second. I think everyone is a data person, whether in your lives, on your phones, on your work. It's just at different levels in terms of technical capability, right? So democratization of data Correct, right? So I think, from an AI perspective and I'm talking about generative AI capabilities even data in organizations organizations are saying actually, all employees, to a certain level and a certain degree, need to be able to use these technologies, these processes, these abilities that are there in order to help them in their roles, right? So I think that that's a big chunk of what organizations are talking to us about and that's where we help with things like data literacy, ai literacy, so on and so forth.

12:06
Lots of conversations that I'm in because of my background. The other end of the spectrum to your point is actually, you know what? We've got some really good use cases here. We need people to come in and build us agents so that we can automate these profiles and actually engineer what a good process would look like. So it's the two extremes, and then in the middle is what I spoke about in my intro is human skills, which is actually how do we create that bridge between the two extremes that I've spoken about, where you get really smart data and AI folk who are able to talk the business language, and vice versa. Yes, and that's been a big part of what we've been doing as well, and a big thing that has been really popular with our clients around that as well.

12:46 - Mike (Host)
So let me jump into a rabbit hole that we haven't prepared. Because of your background, this will be a walk in the park.

12:53 - Waseem (Guest)
So, because of your background, this will be a walk in the park.

12:58 - Mike (Host)
So I was reading the other day about my natural brain as a process consultant by background is we're going to use enabling technologies, generative AI, to make existing processes more efficient. But first I'm going to look at what the business goals are. I'm going to do the right thing think about business goals, think about strategy, think about where the biggest opportunity is, and I'm going to try and automate that. And then look at some things where I go.

13:24
Actually, we should stop doing that stuff. It adds no value and we should do more of this stuff over here because it adds more value. And in the stuff that adds more value and the things that we should automate, I'm going to look at the business process. I'm going to do a business process mapping exercise and then I'm going to use people like you to build AI agents that can automate all of that process in a linear kind of fashion. Is that thinking flawed because AI is different? Or is that what you do? Do you literally do a process map and say now I'm going to use an N8n type environment to automate that workflow?

14:01 - Waseem (Guest)
Yeah, I think there's isms to it. So high level automation has been around for years, right? Exactly? Yeah, nothing new, is it? Machine learning, it's been going on for years, correct, and even basic automation in it. Do you remember Microsoft Access and VBA? Exactly.

14:19 - Mike (Host)
It's been happening for ages, right, so there's nothing different in the core principle.

14:24 - Waseem (Guest)
The technologies have evolved, of course, but there's nothing different in the core principle. So, at a high level, do a process map. This is the process. Here's elements that we can automate. Here's elements that we can automate. Here's elements that we can't. Let's pull something together in order to do it. The isms to it are two things. Right. How do you bring the business along with the journey, on the journey of this new technology? Because there is a little bit of fear around AI because of media hypes and so on and so forth. So how do you ensure, culturally, you are bringing the business on the journey in terms of how to actually manage this going forward? And you and I will both know this, mike, that actually, if you don't do that bit correctly, you could bring the best technology in the world, exactly, completely irrelevant.

15:08
So I think that ism is really important. So we pride ourselves in taking that into account whenever we're dealing with projects like this and working with clients on that and almost challenging them like a partner would in terms of how do you take this into account? The other ism around generative AI capabilities is ensuring that it's not what you're using in order to create the agents, what you're using in terms of large language models. It's understood that it's an augmentation of what a human would do in the example of a human automation process, right. So let's think about marking and it's at the top of my head at the moment, because I was talking to a professor about this right, and actually a professor of university, and they were talking about how it's a big debate around whether they could automate marking and there's an ethical consideration around whether they can automate marking and there's an ethical consideration around whether they can automate marking using general AI capability.

15:57 - Mike (Host)
So for the listeners, this is about students writing exam papers and using AI on the marking side to mark the homework effectively.

16:08 - Waseem (Guest)
Or the other way around, right, or the other way around exactly. Probably more on the project side as opposed to the exam side. Right, because I think, with certain regulations around the exam side, not not not an expert in this, but but more on the project side. And actually we were talking about it from the the marking perspective and the student perspective. Right, the reality in in that situation is you wouldn't. As a professor. So I was like would you just use ai to just say, okay, right, I don't know, you're marking a database diagram as an example. Let's say that project. Right, would you pump it into ChachiBT or equivalent and just take that verbatim? And he's like, no, of course not. I'll look at it and I'll just make sure it feels right, but it will speed up the process for me. And I was just like, okay, that's good, right.

16:50 - Mike (Host)
So that's human in the lead marking.

16:52 - Waseem (Guest)
Using AI as a more efficient way to get through the volume, spot on, and it's the key for me, in what Augmented, exactly right, and even in like. Let's talk about when you and I went to uni, right? So whenever I had a project, I would use a book. I would open a book, I'd look at book, I'd look at resources, look at book, I'd look at raw resources, I'd reference those resources. Maybe people slightly younger than us used google, but now people are using generative ai and, as long as they're using as an augmentation, as a resource, and understand the flaws and so on with it and hallucinations and all of those kind of that's the discussion happening at the moment. So so they're the two isms in in the scenario of your process map. Let's get the business on board and understand. This is an augmentation as opposed to like completely. Let's replace a replacement. Yeah, and there will be certain things. And just a caveat here right, because there will be a challenge here and it's the right challenge.

17:42
There'll be certain processes that could be fully automated, right, but my challenge is they've existed for years. There's been years and years that they've existed in organizations and you can go in and you can write a bit of code and you can fully automate the process. Very easy right. The expense process in organizations used to be receipts. You used to give it to someone. They used to span it in. Now you could take a picture of your receipt and it will automatically do it. Simple right. So I think there's isms to that, but at a high level that process has existed.

18:09 - Mike (Host)
So let's again resourcing. So, having been in the resourcing world, big companies deploy hundreds or thousands of contractors and the traditional model is you find very skilled contractors that have a very particular experience base it could be project managers, it could be data scientists and you deploy those into an organization. But once you deploy them, the organization that finds the contractor effectively lets go. They give them to the organization unless it's a statement of work or deliverable-based, but if it's just a contractor, they're then really the responsibility of the company that's using them. Obviously anyone listening within the realms of IR35, outside IR35, all that kind of good stuff but there was no training of that contractor done.

19:02
What I'm hearing from what you've said which I think is a really interesting model, and I'm seeing a few others doing the same is actually as a big brand using contractors. If we had a rule that said or a principle, all our contractor resources had to be AI literate and had to bring tools with them, all use our tool set, be trained in our toolate, and had to bring tools with them or use our toolset, be trained in our toolset that would make life more efficient, a lot more efficient. Is that where the market's going? Is it now is there much more emphasis on you as a sourcer of talent, to upskill that talent before deployment. I guess is what I'm saying talent?

19:42 - Waseem (Guest)
to upskill that talent before deployment. I guess is what I'm saying. So the Rockball model is probably closely linked to the contractor model in recruitment and talent sourcing. But the difference of the Rockball model is that there are employees effectively right, so we employ them, we'll train them, we'll develop them.

19:56 - Mike (Host)
Ah, so that is different.

19:57 - Waseem (Guest)
And then we'll deploy them, and we'll deploy them on a rate, but we will upskill them based on client demands. So you're the employer of record Correct right and throughout their time with the client, we will upskill and train and develop them and so on and so forth. There's an ism to our model where they can hire them in after two years and they can make them a permanent number of staff. But because we are owned by and we're part of a group which has a big, large recruitment brand on it, I can talk a little bit to the contract side as well. I think what organizations are looking for is people that they can trust to come in and deliver something as a true expert in the contractor space right, and who has got the expertise in order to come deliver it.

20:37
And actually what we try and do and because of my background we've done this a few times now where we actually bring some of our contractors onto a training course that we'll be delivering and actually take them through that training and say actually you know what? We've got AI applications right. We ran I think it was about a year ago now we ran a big webinar for AI applications and actually I was like why don't we bring our contractors on board as well? Because this will be really interesting for them. Now, why is that good for us as a business?

21:01
Those contractors would like to work with us because they're getting opportunities. Why is it good for the client? Because we're upskilling people that they're using in their business at the same time as well. So I think I couldn't possibly comment on whether the whole contractor market is going towards that, but I think there is a genuine appreciation from the client side and the contractor side when the host organization if I can call it that is investing in their development and upskilling so they can then benefit the client and themselves in the future as well. So I think there's definitely a benefit to that, in my view.

21:37 - Mike (Host)
So I think that industry I still keep close tabs on that industry I think it is evolving. I wouldn't say it's transforming overnight. I think that industry I still keep close tabs on that industry I think it is evolving. I wouldn't say it's transforming overnight, I think that's nonsense, but it's certainly. I think there's an evolution about how you solve the problem of I've got skills gaps and how you fix that, and I think that model is changing definitely. So let's now debunk some myths about AI. Business enablement. Lots of talk about huge efficiency improvements yeah, the replacement of humans, data problems being a thing of the past. Can you talk us through the most common use cases that you see when deploying AI and automation, the role of humans, importantly, and the data issue?

22:15 - Waseem (Guest)
Okay, so we probably covered a little bit around the human augmentation side, but maybe I'll start with some of the myths. And actually I've done a talk on this around. I think I called it demystifying something. I didn't come up with the title, but the title was really good. I came up with demystifying AI in some way, shape or form. But actually there is one of the myths that you and I have just spoken about is around AI will replace humans Exactly humans, exactly. That's the big fear. That's the media hype. You know what Some large personalities have come and said.

22:46
Actually, in certain cases, it will do so, right, right. And actually what my challenge is is the reality is AI is being brought in to organizations and actually very few have got it into production. Let me just put that out there as well. Still in the kind of beta phase or the BOC phase is an example, proof of concept phase, but but actually the reality is it's being bought into, all about someone's role and making better. So what individuals need to do is learn to upskill in these technologies, because actually their role might broaden as a result of bringing these new technologies into into play, which is why we'll be in so much demand for people around AI, literacy and so on. Right? Is there certain roles that may go as a result of AI? Yes, but that existed for years, right yeah?

23:29 - Mike (Host)
When we had the Industrial Revolution. That was why people walked down the streets saying this is the end of humanity. Effectively, and it wasn't we evolved.

23:40 - Waseem (Guest)
Exactly right. Even going back to the farmers and picking vegetables and all of that kind of stuff. That's right.

23:45
Like machinery came in, it was like, oh God, what's going to happen to the farmers? They learned how to use the machinery and they could do it faster and better, and so on and so forth. So I think that's probably the first one that I would want to talk about. Right, and I'm not sitting here saying absolutely AI will not take jobs or anything like that. All I'm challenging is that's existed for a while and it has been through various different phases. Right, and if organizations are looking to cut costs, they might use AI as an excuse, but they probably would have done it if this technology hadn't come about either, because there was economical strains on the business and so on and so forth.

24:16 - Mike (Host)
Thanks for listening to Higgle the B2B sales club podcast series with your host, mike Lander. Please Higgle the B2B Sales Club podcast series with your host, mike Lander. Please subscribe so that you'll catch all the next episodes.