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.
Amel Edmond (00:00):
We actually have a very strategic mindset towards artificial intelligence. It's really from two different angles. It's top down and bottom up. Leadership sees the value in artificial intelligence, but we're always cautiously optimistic on the actual impact of it. We need to make sure that we have APIs around it so that we can measure the impact. The good news is our innovation team has a great ROI calculator as well as PPI indicator that tells you exactly how our artificial intelligence has impacted our organization.
Jon Herstein (00:30):
This is the AI first podcast hosted by me, John Stein, 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. My guest today is Amel Edmond, the CIO of Whitham, Smith and Brown, and I'm your host as always, John Hursti, chief Customer Officer At Box. Every episode we talk with CIOs, IT leaders and AI pioneers to share their real world insights that you can act on today. And with that, let's dive into today's conversation. Welcome, Amal. Thank
Amel Edmond (01:11):
You for having me. John.
Jon Herstein (01:11):
It is my pleasure. Can we start with a bit of background on Ham Smith and Brown and then a bit about your role in Koko, your role over there?
Amel Edmond (01:20):
Certainly, Woodham Smith and Brown is just over 50 years old. The accounting firm has actually stretched its arms around advisory services and we do a lot of obviously audit, tax, those sort of things, but advisory has really been an interesting sweet spot for us. Again, just a little over 50 years old. We've grown drastically over the last 10 years. In my tenure I arrived, the head count was approximately 750 users. Currently we have just over 3000. A lot of that growth has been over the last couple of years because of just what's happening in the industry around Harvard equity. There's other variables in the accounting space, but again, very exciting firm to be a part of. I originally came from eiser, came here as the head of security and infrastructure and graduated to the CIO role. My team is very, very diverse. We actually pride ourselves on being able to manage just about everything in-house, whether it is cybersecurity or app development and even operational support. All of that is being handled in Charlie from a very, very lean and super productive team.
Jon Herstein (02:37):
That's incredible. And I think you account yourselves among the top 25 US accounting firms. Is that right?
Amel Edmond (02:43):
That is correct. Top 25 firm. Still vastly growing, trying to break that top 20 barrier, but I believe we'll get there still.
Jon Herstein (02:52):
Can you tell us a bit more about the advisory services line? You said that seems to be a faster growing portion of what you do and maybe a bit different than typical accounting. What exactly is included in that?
Amel Edmond (03:03):
Yeah, so advisory services really encompasses a lot of the technical aspects of advisory, so we have separate orgs underneath that practice. That practice is actually led by very, very smart. Jim Burke, who's pretty much a rock star in the accounting industry. He heads up our cybersecurity team. We have a digital transformation team, we have an artificial intelligence team. We even have a healthcare business intelligence team, so on top of that, there's also SOC artists that we perform. We perform PCI surveys, Oxley. There's nothing that this team can't do and they're led by very, very smart individuals who are really laser focused on customer service.
Jon Herstein (03:47):
Incredible. What I want to kind of pivot to now is a bit about you and we've talked a lot about your fascination with how things work and curiosity about the world, both technical and non-technical, but I'm wondering what is your personal why for investing and making the effort around AI today?
Amel Edmond (04:05):
AI we all know has different facets to it. There's generative ai, there's regular ai, there's all these different flavors to it. For me, it's technology. It's technology as a whole. I pride myself on not being a CIO, but more so as a technologist and I think the rest of my team follows suit Technology is the way that I define it as someone who has a passion for it, someone who would be doing this even if they were not getting paid for it. Ultimately it's nice to be paid for it, but ultimately, again, it's about the passion for what you're doing. Now, to answer your question around artificial intelligence, it's really the biggest craze, but everyone knows that it's been in the works for about 40 years if not more.
(04:51):
It is really probably the biggest change in technology that I've seen really taking over globally. I believe in the last two or three years there've been more AI-based companies starting up in selling and just really being very, very profitable in the last three years than any other type in technology history outside of the dot. So that's why I think that again, AI is definitely going to be revolutionary. It continues to play an impact across the board, whether it is counting finance, even military applications, AI has been there and right now it's become way more mainstream. Even the time it takes to develop applications or even code checking for example, and it has been drastically reduced and the impact on our user base or resources has been drastic.
Jon Herstein (05:46):
I want to talk about the application of AI and how you're putting AI into practice, and I think you've been at this for some time now. One of the things that we talked about earlier was a reporting automation project that you worked on around 10 99, and I'm sort of curious if you could give us a bit of background on what that effort was all about, what the problem was, how AI helped you solve it, and what are some of the outcomes that you see and as a result of it?
Amel Edmond (06:11):
Yeah, without spilling the secret sauce of
(06:15):
10 99 process, we all know that it can be very arduous and if you actually, before I even dive into that, let's go back for a second to understand the need or the urgency around artificial intelligence specifically to accounting. There are much less people graduating out of college historically with CPAs or with accounting degrees in general than have been historically. What that does for us is it creates a resource shortage. Obviously we can look at resource options internationally, which we have. We actually have a great team led by an international leader out in India that actually spent up two offices currently out in the India space. But the problem here is the work will still be there. So the advent of artificial intelligence brings about a great opportunity where we can look at certain processes and attacks or audit or even advisory space to try to streamline that and really empower our team, not so much to constantly be focusing on the actual task or checking off that box or going into an Excel spreadsheet and ensuring all the tasks are being executed effectively. That is something that can be done via RPA robotic process automation or artificial intelligence as we continue to leverage this. Now,
(07:36):
How are we're doing that in the sense of our 10 99 or 10 40, all these other IRS codes, we're literally documenting and understanding what our processes are, and then the issue that we always find is processes get tweaked based on leadership and location. The first step is always to standardize that process. Once we have that process standardized, in the case of this, and in this case it's turned 99, once we have that process completely standardized, we try to inject automation and portions of that process, whether it is by creating an entire different workflow or adding some sort of script here and there. That's how we started down on this journey, but ultimately we started to make these scripts a lot larger and really compressed down the process of reviewing that 10 99 and once we were comfortable with the process being, with the answer of the process being accurate, we started to allow to expand even further. Again, that 10 99 process is fairly arduous at times, but the goal ultimately is to again, free up our team to actually be on the process and not the task up.
Jon Herstein (08:49):
Yeah. Something you mentioned in there that has really struck me in the last few weeks of conversations with customers is that everyone seems to start with making a piece of the process more efficient, maybe faster, maybe automate that piece, but what's maybe taking a bit longer is to take a big step back and look at the entire process and think about how you reached that whole process given these capabilities, and I think it's something that we could touch on a little bit later in the conversation. You also mentioned a cloud-based audit platform that you've been working on that combines the power of some of the newer AI technologies, but also machine learning. And let's not forget about machine learning as being a really important enabler of some of these kinds of capabilities today. What are some of the biggest changes you've seen as a result of the combination of these capabilities and are there any specific metrics that you found surprising in terms of how you can improve given these capabilities?
Amel Edmond (09:44):
Yeah, so our audit team has been one of the first, actually if not the first, to actually complete an entire audit from beginning to end or the new CaseWare product called the Dynamic Audit Solution. That platform has been fairly revolutionary because when we looked at the industry maybe seven or eight years ago, firm leadership decided it was worthwhile to start investing in heavy artificial intelligence, but something that had that built in from the ground up incomes, CaseWare were actually proposed a very good option for us, and then we went down that avenue along with them. Two years ago, we actually started that process on testing exactly how it's going to impact our organization specifically on the audit side. From there, we actually read through an entire audit and it was very successful and we started that process of transitioning away from the existing platform to the need platform, and what it actually does is it actually helps you streamline a lot of these different processes and again, this is all about saving time and for us, if you back away from the table a little bit without looking at all the melts and bolts, we back away from the table.
(11:00):
Ultimately, how does a firm light hour break into the top 20? We have to be more competitive. In order to be more competitive. Sometimes it's about providing the right amount of the right technology. On the other side, it might be the business side. We can actually just be more competitive pricewise, so if you're actually shrink down the time of engagement, if you're actually able to shrink down the cost to your customer, ultimately you're able to produce the same amount of work, if not better because of the consistency and less human errors as well as you can reduce the price of the engagement, and that's our go to market play.
Jon Herstein (11:34):
Right, and that's a great point. It's not just about doing things faster but actually with higher quality at the same time, and normally those things essentially compete with each other. You're able to do both.
Amel Edmond (11:43):
Correct.
Jon Herstein (11:44):
One other project that we talked about was your work with one of our great partners, Varonis, and maybe you could talk a bit about what you've done there, how that's helped you make all the content you have in Box even more secure than it was already?
Amel Edmond (11:56):
Absolutely. First and foremost, you have to understand that the majority, if not all of our information is actually sitting in Box, specifically all of our client information that allowed us to share and collaborate with our clients very, very effectively. Now, layering in vir, Varonis was a no brainer. Varonis allows us to now reduce the amount of exposure that we have. For example, if a new person comes in and that person works for let's say the finance team or the HR team, they don't necessarily need access to the entire client directory, and Varonis will allow us to do just that data segmentation as well as data security as well as allow us to understand fully exactly where all of our data is, whether it is on the box ecosystem or the Microsoft ecosystem. So having that single pane of glass for governance is going to be a good game changer. It's currently is still being released, but we should have it completely in production by the end of this month.
Jon Herstein (12:52):
Alright, well I will check back with you in November and we'll see how things are going.
Amel Edmond (12:58):
Absolutely.
Jon Herstein (12:58):
So I'd love to pivot back to AI and maybe step out a little bit and talk about how you think AI will start to allow your practitioners to surface insights from all of the client content that you have. You mentioned all of your contents in Box. Obviously we've got capabilities for AI on top of all that content, but how are you starting to think about making use of those capabilities from AI combined with all the content you have? Are there early ideas or use cases you're already exploring in this realm?
Amel Edmond (13:33):
Yeah, so what great use cases with our legal team. Again, very, very strong team, but again, very lean. So how does a team of that size manage all the different agreements and engagement letters that are coming in on almost a regular basis? Daily, they look through 50 contracts almost a day. How do you summarize one versus the other and enter box ai, we can literally have all of all those contracts in one repository. They can actually do a comparative between all of those and actually give summaries relatively quickly. Some of these contracts, believe it or not, can be hundreds of pages and right now our legal team is able to look through a contract, get an accurate summary of that file and even do a comparative between one version and another. That's one of the major ways that we have immediate impact from let's say boss AI or AI in general. You can do some of the similar things in copilot for files that are sitting in the Microsoft ecosystem. Some people are using that, but there's always that hesitancy to leverage copilot just because of the ship fact that you are not a hundred percent comfortable with having all of your information in that Microsoft ecosystem unless you're certain that the data had been segmented and the proper layers of security have been put in place. Thus, like I said before, earlier, having a platform like Box or Varonis as a data governance layer helps make that possible.
Jon Herstein (14:58):
Yeah, what's nice about that is you've already agreed to a set of policies and a hierarchy for who's got access to what, and that's already instantiated inside of your box instance and now you're just bringing AI to that already existing structure. Exactly. Yeah. So you talked about some of the internal processes that are being improved. What about on the external side? So if you think about things like advisory services, are you starting to use AI at all to help improve some of the deliverables, make analysis faster, richer insights for clients? How is AI starting to affect your customer facing work?
Amel Edmond (15:32):
Well, the biggest customer facing product that we're working on right now actually is being handled by our innovation team. Great team out there, heavy duty app development and definitely focused 100% on artificial intelligence. They're working on something that's really revolutionary, should be ready for full deployment by the end of December of this year, it is going to be a client experience portal. What that does is it's a AI based solution where the client can go into a portal, access all of their information, upload files, have communication or collaboration with the people that are actually involved in their respective jobs or engagement and have full visibility. So no more CL and Dagger, no more trying to figure out exactly, Hey, can I get an itemized bill? How much time did a MEL spend on my job? Everything is there full disclosure, and again, that's part of our go to market strategy to make us fully transparent to our clients and allow them to be a part of that journey for their engagement.
Jon Herstein (16:31):
I love that idea of a client facing experience portal. That's AI first. Every interaction is going to be powered that way. Exactly. And it's got all the context of the relationship you have with them, all the deliverables, the documents, everything is all right there and interestingly, we're thinking about a very similar thing for our own sort of digital front door into our support organization. So maybe we can compare notes on that after the podcast.
Amel Edmond (16:56):
Absolutely, and don't forget, again, I love picking on my legal team because we work with that so closely. We're also working with Box on an AI based contract management portal with your team, so hopefully more to come from that very, very soon.
Jon Herstein (17:13):
Yeah, I would definitely love to talk more about that. Maybe it's a little bit too soon, but can you give us a sense of just what is the vision for this?
Amel Edmond (17:21):
Yeah, the vision is a master repository for all contracts globally and being able to instead of on file by file and folder by folder basis, you can literally have a full visibility of every contract cost per for license. Many people that's involved that it's actually going to be the contractor is going to be terminated, who was involved in the execution signatures, you literally have all of that in one single pane of glass. Obviously the repositories, one aspect of it, the AI on the backend will be able to extrapolate that information and present it to you in whatever power BI format or dashboard works, but you can actually interact from the dashboard to the live document interact. And on top of that, the other goal here is to have multiple people be able to interactable with a file accordingly. So if you have a senior partner that's looking through an agreement, and that's something that's still missing from an IT perspective, you can literally say, wait a minute, this vendor doesn't have a SOC two type two in your repository. These are your minimum qualifications. This even no can sign this agreement this way. Nothing slips through the cracks, nothing goes around our vetting process. It all ties all into that platform.
Jon Herstein (18:33):
One of the things that I'm seeing start to kind of pop up more and more is this idea of multiple very task specific agents looking at the same underlying material with very different intent. So if you take that example, maybe you've got a compliance agent that's looking at that content, those contracts, et cetera, but only looking at it from a compliance lens and surfacing issues. That may be a concern there. Maybe there's a completely separate commercial agent that's looking at it from the standpoint of is this going to be a profitable contract for us? And all of them can be working simultaneously on the exact same content, which is I think a pretty powerful thing.
Amel Edmond (19:08):
And something like that would tie right into something like a trust center, which was another piece of technology that we have AI based reviews contracts and you actually upload all your questionnaires into it. Did it autonomously answer the questions of the questionnaire based on previous answers as well as based on written policy that we currently have in place. Again, a lot of implications of AI, at least in our environment, and I know that they can be great across the board for others.
Jon Herstein (19:31):
Yeah, I've seen a very similar set of use cases around RFP responses or grant writing responses, very, very similar kinds of workflows. So it's great to see you working on that. You mentioned the innovation team. It's clear there's a lot of innovation happening. I guess I'm curious whether at with that happens naturally and organically or there are things that you and the leadership team are doing very specifically to encourage that. What's from a cultural perspective, how does this actually happen?
Amel Edmond (19:58):
Yeah, so we actually have a very strategic mindset towards artificial intelligence and it's really from two different angles. It's top down and bottom up leadership sees the value in artificial intelligence, but we're always cautiously optimistic on the actual impact of it. We need to make sure that we have KPIs around it so that we can measure the impact. The good news is our innovation team has a great ROI calculator as well as KPI indicator that tells you exactly how our artificial intelligence has impacted our organization. And truth be told, over the last four years of actually three and a half years, our innovation team has saved us over $15 million because of their automation and their artificial intelligence. So hats off to that team. They've done an absolutely monumental job for the last three years, but there's a lot of different aspects of artificial intelligence that also could be done on another side.
(20:49):
For the coin security, we're running a very, very lean security team here with IT department, but we're not necessarily a small team as far as our impact small team, but a lot of technology, a lot of artificial intelligence we're using, again, can't tip my hand too much, obviously this will be aired somewhere, but we're using artificial intelligence on basically every aspect of our environment, whether it's the networks, the endpoints by endpoints on laptops, whether it is on user itself as well as our data in our data repository. So artificial intelligence has pretty much baked into every aspect of our environment, whether it's next generation firewalls, our networking or our actual connectivity between the different sites.
Jon Herstein (21:37):
That's incredible. And you'll be happy to know, we announced actually box force just a few weeks ago using similar capabilities and building it into our SHIELD Pro product and basically saying AI can help make box even more secure than it is now by doing the same, the kinds of behavioral analysis that you're sort of referring to there. So I think it's happening everywhere and we're definitely happy to be a part of that change with innovation and with cost savings, you always have the risk or the fear of automation replacing jobs, replacing roles, maybe even just changing the roles that people have. So I'm sort of curious in your industry and maybe with 'EM specifically, but I just think maybe more broadly, how is that being received by people who are doing those jobs for a living? Are they concerned and what has leadership been doing to help build trust in the future here
Amel Edmond (22:31):
From a leadership perspective, the message comes directly from the top. Our CEO, pat Walsh has been really focused on making a direct message during our town hall saying, listen, AI is not here to replace you. And if you remember earlier in our conversation I talked about the shortage of accounting majors that are graduating now, the whole point of this is to really empower our people. We're not looking to get rid of anyone, we're not looking to do any things of that nature. My goal is to provide the right amount of technology to make sure that it's impacting our firm effectively. Leadership's goal is to make sure that our people are working, are being engaged and are still focused on providing the best client service that they possibly can. And to do that and have to have the right technology, you have to have the right AI in place.
(23:23):
So again, it's a two-pronged attack leadership message as well as bottle up. The good news is newer people coming in, newer, I'm going to say younger, newer staffers that are coming in, they're already looking for that. Not having artificial intelligence as part of your technology stack actually turns younger people away. It actually attracts them. If you actually have it as part of your marketing, as part of your onboarding, say, listen, we've got a bot that can answer your question. They actually enjoy that because guess what? I put a tablet in a child's hands and that 2-year-old was able to swipe left and right. Many moves ago, I'm pretty sure if you put a tablet in my hand, I would know what to do with it except for breaking it. So it is a different generation, so we have to actually be willing to lean into that generation and leverage them more effectively.
(24:11):
And what that does is unfortunately there are going to always be some individual that are resistant to change. That's why you attack it from the bottom up, gamification, create some level of fear of missing out. Hey listening, Jane got a $50 award for coming up with this great idea. Let everybody else get involved with it. And after a while it becomes infectious and continues to grow. And that's the mindset that we have here with up making sure that our people are engaged, making sure that they're all excited to be here and a happy person means great client experience. And at the end of the day, that's what we're all about.
Jon Herstein (24:48):
And in an industry, as you said, where you have the challenge of maybe not enough people coming up through in the early stages of their career, so you actually have a labor issue to begin with. So that probably makes it a bit less of a scary thing for people to say, actually, I could use the help because there's not enough of us doing this work. Exactly, yeah. Have you all started talking about what all this means in terms of not just how many of these folks there are, but actually the nature of their role? How does the role of an accountant or an analyst or knowledge manager change in the age of ai?
Amel Edmond (25:21):
Right, so what it means is a staff Y would now actually be pushed up to the level of staff too. The responsibilities will still be there. So if you automate a process, there's still going to need to be a human to review it. There still needs to be some level of trainings for that. The staff one understands the why that we're doing this, simply clicking off our checkbox or any bot can do that. So the understanding is why are we looking at this? And what that does is from the top down, it frees up partners from actually helping or and doing more of the work, pushes that work down to the senior managers and manager level and allows the partners in our organization to go out and become business advisors even more, allow them to actually do more business development work from the bottom up.
(26:08):
Now you're looking at potentially removing one layer from the traditional work to allow the staff ones to do more. Staff two, work more responsibility means you have better exposure. All too often we've heard from staff ones, this is kind of mundane, I am not learning anything. So now with ai, you're owning a process, you're owning the actual totality of a process versus just checking off a box or reviewing documents or doing research. All too often we've seen research, for example, as another item for artificial intelligence. Research can take sometimes days depending on the engagement of what you're trying to accomplish. Now you can actually get research done in a matter of maybe an hour or two and it's just a matter of validating it versus spending three days when you're in queue researching a particular case.
Jon Herstein (26:59):
I think this concept is so important of elevating the roles that people have and taking out the kind of work that mostly people don't really enjoy doing. To your point, the level one folks are doing work that they don't find that interesting that they consider rote repetitive, not that challenging maybe and maybe they're not learning a lot from it, but it's not so much that we don't need them anymore, it's just that we can now redeploy them on the level two tasks or the level three tasks and everyone sort of shifts up. And I think that's the vision for how we leverage these things
Amel Edmond (27:32):
And that should definitely be the vision across the board. Not to replace but to re-skill.
Jon Herstein (27:38):
Have you started to think about not just AI augmenting the work of individual folks, but actually agents that are taking on some specific part of the process completely maybe managed by people or how are you beginning to think about agents in the workforce?
Amel Edmond (27:52):
For me, it goes back to how the IT department handles artificial intelligence and agents in general. Ultimately, I don't want an agent making critical decisions like shutting off a server, but there needs to be multiple alerts there. Shutting off a task, for example, or restarting a service that's pretty much standard. That's an auto restart task that could get, can be done in a one on any old school Windows servers. Now imagine taking that to the next tier. You're seeing, for example, me, let's say amel is normally working in it, but suddenly he starts downloading all of these files out of the HR directory or looking into client directories where he normally shouldn't. There should be an alert that says, let's put this under a risk management protocol in a sense, let's start monitoring this to see what's going on, start downloading files to see if there's anything nefarious happening here. If everything clears, it checks off by itself, but a report still goes somewhat to validate that everything was clear or not. So that's the way that we do it here at wood.
Jon Herstein (28:52):
I want to pivot a little bit to you personally and just as a leader who is embracing this innovation, embracing these technologies and capabilities, are there specific things you would say you do differently now than you did before ai? I mean it's only three years ago now that some of these capabilities came on the scene, but what's changed about your leadership approach because of these capabilities? Now
Amel Edmond (29:14):
My leadership approach really hasn't changed too much. We're always looking for the next great thing that's out there. So we're always looking at evaluating technology across the board, and that's been the methodology behind the entire firm. Leadership has always invested strongly in technology, which is while we are where we are, but me personally, the only thing that's really changed is how I look at defending against AI based attacks. So you have to have artificial intelligence to combat against artificial intelligence. And believe it or not, bad guys are out there and they're leveraging this very, very well, and they're actually probably leveraging it more efficiently than most people because there are no rules and regulations around 'em. So again, leveraging that and creating the proper environment that you can understand and learn from the bad guys by. I say that by setting up my honeypots and things of that nature, or even if you're a strong enough technologist, you can go on the dark web and actually have these conversations with these individuals and they'll actually show you how they're breaking into things and all the other things.
(30:15):
I actually spent some time maybe about a year and a half ago on the dark web where a group of individuals actually created a playbook on how to break in with account fairs. They had dates, deadlines, how to actually break in, because obviously during deadline time, accountants are laser focused on getting the engagements out, but guess what? The bad guys also know that you're under pressure. So guess what? You suddenly get a mystery little popup saying, Hey, by the way, this is the help desk. You just click on this link and uninstall this one file. And before you know it, you have someone in your environment potentially trying to jump laterally to break into other systems.
Jon Herstein (30:51):
That is scary, and I'm glad you're paying attention to it from the defensive side of it. And you make a great point about the good guys. All of us are thinking about all the ways that we need to make sure we're compliant, that we're still following all the rules, we're taking advantage of ai, but not doing it in irresponsible way. The bad guys don't have to think about that, right? So they've got a bit of an advantage that way. So you're in an industry that's quite conservative, and it feels like today you can say that the firm's really embracing all of this, but was it always true? Did you have work to do early on in a pretty risk averse industry to try and convince the various stakeholders to take on AI and embrace this?
Amel Edmond (31:34):
Yeah. I'm actually in a very unique and blessed position where I got into this firm and they were saying, what do you need to make us successful? So I didn't have to come up with a very long 30 page business case. It was simple conversation where they trusted and believed in my vision and I said, we need to have these specific components in place. Let's look at it from analyzing your current environment, coming up with a game plan and understanding where you need to be. And I align my vision with the furnace vision to actually move forward, and I gave them the bill of whatever millions of dollars it was, and guess what they said? Yes, let's go ahead with it. It wasn't a very long conversation, surprisingly. So again, not all firms, not all organizations are built that way. There's usually a slightly longer process, but my firm was actually really focused on getting where they needed to be and trusting the process and trusting the team members that worked for me.
Jon Herstein (32:35):
That is a great place to be. You picked a great firm and it's going well. Now as you drive all this innovation, you obviously still have to be compliant with all sorts of regulations in both your industry with the sensitivity of your client data that you're dealing with and so forth. So how did you find that balance? Did you take responsibility for that? Is that another team that's looking at that? How did you make that all work?
Amel Edmond (32:59):
One of the things about my position is I actually wear multiple hats, not the chief of information officer position, but ultimately everyone understood that my passion was always in security and keeping everything safe and trying to make sure that I understand how everything works. So to answer your question, yes, ensuring that compliance is where it should be and we're following the proper regulations and we don't end up in the news that follows on my responsibility. But we recently hired some of actually as head of our GRT government risk and compliance. He's in charge of reviewing everything, whether it's policy data repositories, designs to ensure that our practice is actually meeting the regulatory standards that were there. Good news for us so far, everything's been thumbs up. The only big issue obviously, are new firms that are coming in. They're not always as up to date as we are. Sometimes they're actually advanced, but in some cases you have to look to see where they have perhaps HIPAA data or maybe something is not follow a retention policy, things of that nature. That person's responsible for doing that. And hopefully that's one thing that I don't have to stay up till two o'clock in the morning doing.
Jon Herstein (34:14):
It sounds like you stay up till 2:00 AM doing other things anyway, so just not that. Just out of curiosity, have you published, written and published any sort of AI principles or AI guidelines that are followed and what does that look like?
Amel Edmond (34:28):
Yeah, that's part of our acceptable use policy, which is part of our handbook. Our team members all have the handbook and it's actually posted on our internet site, which is actually, you can query it against our internal built ai, we call it ivan. After our founder, you can actually ask a question of what's our policy on mobile devices? It'll look up the employee handbook and give out the answer. So yes, the short answer is yes, we do have our AI policy common and don'ts. The long version of that is all sitting in our acceptable use policy, which is underneath our employee handbook
Jon Herstein (35:04):
And all powered by an AI chat bot interface sounds like.
Amel Edmond (35:08):
Exactly.
Jon Herstein (35:09):
So you can be as vague or as specific as you like on this next question, but I want to ask you, in the spirit of experimentation and people testing things out, sometimes things don't always go as well as you'd like. Things fail. We have a value at box about failing fast. Can you share a project or an initiative that you started that didn't go as planned and maybe you had to abandon or change course, and what lessons did you learn from that?
Amel Edmond (35:36):
More recently, I can tell you when that was very, very dear to my heart. We designed a reporting platform that essentially would allow our team members to understand where their report card is essentially. So imagine a platform where you can look to see how many days you've been in the office based on obviously firm leadership saying, let's get back into the office for X amount of days. You can look to see how many days you're in the office, how many billable hours that you've actually put in so far, where are you in reference to your total number that you're supposed to meet, all kinds of other metrics that are pretty cool in there. And it literally pops up on your screen every day. You look at it, you hit acknowledged you have a digital signature and you're done. That was the great idea behind it, of course, firm leadership, they said it's a great idea, rolled it out, made an actual application out of, it wasn't really, it was kind of clunky, would have issues with failures.
(36:31):
Sometimes it would pop up and freeze your entire computer. So that's our drawing board. The design at its core is still there. We are still keeping that design in place, but we're actually making it more embedded into the web browser now. So hopefully in the next two to three months, that platform will be finalized and it'll be web-based. So now you're not looking at another application on the endpoint to use up more ran or you're not looking at a platform that needs to always be updated. It's part of the Microsoft ecosystem. It leverages edge and that's the goal. So that's one of those classic examples of what we should have slowed down and not rushed to design something. We should have actually looked at the totality of the agreement and said, alright, let's not make it an application. Let's put it here or here. There should have been other options. But again, sometimes things fail, but you fail fast, you fail slow sometimes. Unfortunately in this particular case, you fail fast and you try and make the best of it and learn from those mistakes.
Jon Herstein (37:32):
Well, and one thing that I've seen is if you restart an effort like that today, and let's say you first did it six months or a year ago, the way you would approach it today is actually probably pretty different than the way you would have a year ago because of the capabilities that are available now, coding tools and et cetera. So you probably do it better the second time than the first time.
Amel Edmond (37:49):
And that's the great part about being innovative or being part of an innovative firm. There's room for mistakes. It is not a situation where you made a mistake and you're not allowed to touch X, y, Z product ever again. It's all right, great. What have we learned from here? So now we know not to do this. Let's take a full review of everything else that's out there today versus making decision based off of legacy information. For example, Hey, I'm going to look at Chad GBT for everything. No, there's other all leverage models out there. There's other GBT type platforms out there that you can actually leverage. And if you actually look at it, they're all specialized. Now, there's one platform I found out maybe a couple of years ago called Tax GBT, amazing platform for research. Again, there's so many specialized items for artificial intelligence. Now the question is which one will be the ring to rule at all? I'm looking for that one. Ag agentic AI or some sort of quantum AI that's going to be able to plug in and rule all of Theis out there.
Jon Herstein (38:49):
And the expression I keep hearing over and over again is these AI capabilities that we have today are the worst they'll ever be, right? It's only going to get better and more powerful. And I think to your point, even more specialized and thinking about how you orchestrate across all these different specialized agents, I think is the next thing that we all need to figure out. And on that note, I want to have you pull out your crystal ball. I want to ask you some forward looking questions a little bit, and let me start with, when you think about all these capabilities, what do you think in terms of what's coming either what's here today and not fully deployed yet, or what's coming with ai, what capabilities do you think are going to most transform your world of professional services?
Amel Edmond (39:32):
If you think about it from an account perspective, it's what's going to affect specific verticals. For example, accounting, or sorry, for example, tax or audit tax has so many different subgroups underneath it. For example, automotive or constructive or hospitality, all of those have specific specialties. So underneath that giant umbrella, there are multitudes of small businesses underneath it. So the question that I always say is, how are we going to manage processes better? Not so much the technology that's associated with them, because from my perspective, software companies will eventually have AI baked into them regardless. Six or seven years ago, AI wasn't necessarily inbox. There was RPA, there was a lot of scripting, there wasn't technically ai. Fast forward to today, AI is baked in. We don't have to invest $10 million and leverage into AI to add to our existing platform. So fast forward four or five, six years from now, my expectation is AI should be baked in. It should be a universal standard. It's kind of like having a SOC two type two for every organization. It's a standard. My expectation is really what I said before. And for one AI to rule 'em all, creating something that can actually govern all other ais, all map out what they actually do, measure out that risk, and as well as measure out the KPIs for each walk.
Jon Herstein (41:03):
That's really interesting to see. Yeah, I think the AI to rule them all, but with specialists in each of these areas, whether the vertical expertise or yeah, I think that's very, very, very likely to kind of play out that way. So when you then think about what the impact of all that is on what a firm looks like in this space, how does it change any aspect of the firm? We talked about some of the roles specifically, but just what does the firm actually look like five or 10 years from now?
Amel Edmond (41:31):
Yeah, so there's a lot of talk right now in accounting about creating a chief AI officer. I've seen that a couple of different locations, slightly larger organizations like Coca-Cola, they can do that. They're 35, 30 6,000 individuals. I'm not sure if that's something that's going to be plausible and an accounting, but it could be. You never know. But I see it as one person that actually will help drive artificial intelligence, adoption, implementation, design and execution top to bottom, whether it is copilot on your endpoint or app, developing something completely revolutionary. So that's probably more a director or a senior vice president depending on the type of how your organization is structured. The way I also see it is that person would actually not necessarily have an army of individuals working for them, but more dotted lines to all the different operational groups, whether it is marketing for distributing information or change management, whether it is it from a security and compliance perspective, whether it is app development team to actually build everything.
(42:36):
Having an enterprise architect or architecture design plan or a coalition of AI people to make sure that the platform is even actually doing what they're actually actually trying to accomplish. So again, there's a lot of different components associated with that, and it all depends on, number one, the company's appetite for that change. Number two, the vision of firm leadership to say this is where we want to be and this is how we want to do it. Having technology is great, but the greatest technology on the planet is nothing but a coaster if it's not being leveraged
Jon Herstein (43:09):
Effectively. So interesting. So sort of a chief AI officer, and I am starting to hear some of those kinds of roles, whether it's called that or something similar. Definitely starting to see that with some of our customers. Open question, what you refer to, does that person have their own team or are they sort of a dotted line across a number of different teams? And there was obviously pros and cons of those two different models. So that remains to be seen a bit. I don't think we entirely know yet, but definitely some change. How about on the revenue side? Do you see a difference in how firms will bill for their services more outcome based? Anything along those lines? Is it too early to predict?
Amel Edmond (43:43):
Yeah, no, actually, I've already had a conversation with Gartner around specifically measuring artificial intelligence and where to invest. Some organizations out there are actually investing 3% of their revenue, 2% of their total revenue in artificial intelligence specifically, not just technology as a whole. It's a very large number in comparison, but again, it all depends on the organization's appetite for that change or the need for artificial intelligence. So in my eyes, I think that it all depends on your organization. Now, the other side of this is there's different flavors of artificial intelligence. Are you leveraging it from an development side? Are you leveraging it from security? Are you leveraging from a utilization perspective? So there's lots of different ways to do that. And all those different measures will actually actually affect different groups or different pockets of your population. Firm admins are generally the most resistant to change because they feel that these bots are here to take over their position, not so much professional services staff. So for me, it's making sure that firm admin understand that this is about maximizing their time so that they're not doing mundane tasks, that they're actually understanding the process. And what that does is ultimately affect profitability to go nothing but up. And at the end of the day, you could run a billion dollar organization with the same headcount as a 700 million organization.
Jon Herstein (45:09):
Makes perfect sense. We've talked a lot about possibilities, but I now want to talk a little bit about limitations and specifically in your industry. Obviously it's a very knowledge centric industry, so AI is incredibly helpful there in terms of just being able to have vast amounts of knowledge, but there's also a lot of human judgment applied in things like advisory services, even a tax strategy for a client particularly, there's not necessarily a right answer, but a bunch of different ways you could do things and human judgment and experience has to come into play there at least today. Do you see any limitations in terms of what AI's role will be in those kinds of things going forward? Will there be a point where you say, yeah, AI is great, but it's not going to do that?
Amel Edmond (45:51):
AI will never be a trusted business partner. Let's be honest. And again, I'm thinking about it from the top down, right? So you can't allow AI to be run your boardroom. That's number one. You can't allow AI to go into someone's business and say, Hey, we noticed you had these three different events to take place over the last three years in your business. If you actually execute this plan or that plan, you can actually be more profitable. That's called being a trusted business partner. AI can't do that, at least not yet. And one of the main limitations of ai, if we actually think about this realistically, AI is really based on data. If you don't have proper data hygiene, AI is not going to work for you. You actually hindering yourself to actually get to your end result by not having proper data. So number one, having a full data assessment or an AI readiness, which will reveal the data issues if you have any, will actually help you understand what that roadmap potentially could be.
(46:58):
So ultimately, for me, it is a matter of understanding the totality of what AI is, what it can do for you, and those are the limitations. To talk about this a little bit more from a resource side, having the right individuals in the right positions is absolutely critical because a CEO's position is not necessarily to make those technology decisions, it's to put the right person in place to help guide those technology decisions. And that would be someone like myself or head of innovation or head of ai. Those individuals should be the trusted individuals. Having the wrong person there could be catastrophic by between the millions or tens of millions of dollars ultimately if you have a long tenure. Now, with that being said, the technology itself, if you have clean data, you actually have the great data governance and you have the proper people around it, and you have a coalition of individuals that's actually focused on leveraging AI effectively, then you have a great recipe for success.
Jon Herstein (47:57):
Very, very helpful. And I wonder if I could ask you for a bit more advice for some of your peers out there as they are thinking about launching new AI initiatives. What have you found would be really important for them to keep in mind in terms of key do's and don'ts when you're starting some new AI initiative? What should you watch out for? Make sure you definitely do, make sure you definitely don't do help out your fellow CIOs and innovators out there.
Amel Edmond (48:26):
Yeah, I think anyone that's currently a CIO knows this already, but we're just going to cover it a little bit. If you're looking for any AI or technology for that matter, you have to really understand what is the goal of your organization, where are you trying to be? And number two, you also have to really be, you have to understand what is the technology going to overall impact, right? Is the change going to adversely affect the user population or maybe your client base, or is this change going to be so painful to convert from the existing platform to the new one? Is that going to effectively have you lose resources? Again, those are decisions that are ultimately left to the business line to decide, but you have to be able to make those arguments before you actually pull the trigger. 'em. Lastly is looking down the line, try to be a little bit forward thinking and think about where this platform is going to be in the future and how it will integrate with all other systems that are already in place.
(49:33):
So streamlining is one of the ways that you can try to reduce overhead. So is the platform in a good situation or a good spot to potentially expand in your environment and take over responsibilities of other platforms, whether it is, let's say practice management for example, or a workflow system that you have place. So you have to be a little bit forward thinking and be careful on how you're actually pushing these technologies out. One of which is for me, it is Workday. We're currently in the process of deploying Workday, and what that's going to do is affect IHRM systems. It's going to affect PSA will affect our practice manager. It's going to affect our entire finance team, and that's a three year journey for us, which is actually a very, very fast roadmap. I've had conversation with other CIOs and other accounting firms, they saved in four or five, sometimes six years to get this sort of change implemented. Again, I equate it to having open heart surgery while running the Boston op. It's very, very difficult for you to rip out your practice management platform while you're a time and billing based organization. So our time is how our revenue stopping that or somehow having any sort of errors in that system could be catastrophic.
Jon Herstein (50:54):
Yeah, you've got to keep doing all of that while you're changing. You hear that analogy of changing the engines while the plane's in flight. I love that. Exactly. Having open heart surgery while you're running a marathon is, I'm going to tuck that one away and use it. Amel. When you're thinking about these initiatives, how do you think about measuring success? And there could be things like accuracy, efficiency, client satisfaction, employee satisfaction, perhaps all of the above. Just what do you zoom right in from a metrics perspective?
Amel Edmond (51:24):
Well, the main thing for us as an organization and we really focus on customer satisfaction, efficiency is great, but having the customer be happy with the end product is paramount for us as an organization. The other thing that we also always focus on is our end users. We try to make sure that we leverage the right technology to balance out their work life balance. Lemme give you an example. Imagine for example, you have a process that takes about a thousand to 2000 clicks to execute, but you can actually apply the correct amount of technology to actually reduce that to perhaps 50, sorry, to perhaps 500 clicks. What that does ultimately is it allows the individual to focus on process. It allows the individual to not have to put in 80 hours a week to execute the same amount of tasks, which ultimately becomes mundane. And what that does has an adverse effect in the person's mind or the person themselves where they tend to want to lead our talent team tracks and makes sure that we monitor and we report on our retention. We are very big on retaining our staff because of the sheer fact that we invest a lot of time, a lot of effort to get that person trained and be comfortable and be part of our ecosystem because ultimately every staff will have the potential to become our next partner. And that's the message that comes directly from our CEO, and it's been that message from all the previous CEOs. So ultimately, again, leveraging the right talent, leveraging the right technology and executing properly is critical for our organization.
Jon Herstein (52:58):
So it sounds like client satisfaction is at the top of the list and the accuracy and the quality of the work that you do for them, employee satisfaction, and then things like efficiency and so forth kind of important too. But those are really paramount, and I love what you said about efficiency is not necessarily a bad thing. You would think in a professional service firm, it's like, oh, we want to maximize billable hours, but the idea that you can be more efficient, charge the customer less for the same service, that actually allows you to increase your revenue by going to getting more clients. And that's I think maybe a counterintuitive way for people to think about it, but I love it,
Amel Edmond (53:35):
Just a different way of thinking. And again, that's what makes weather a little bit different than a lot of other firms out there. We try to take it from a different angle. We try to be a little bit more innovative than our thinking.
Jon Herstein (53:47):
So one of the things I think about a lot in my role is adoption of technology and how do you get end users to actually do something different than they do today? So pretty broad question, but as you think about getting to those outcomes where you are introducing new technology, but you're doing it for a reason to get to one of those things, better client satisfaction, better employee satisfaction, more accurate, more efficient, et cetera, what approaches do you take to actually get adoption of these new tools and technologies across with them?
Amel Edmond (54:18):
Excitement, gamification, all those words that you can possibly think of? We pull every lever we can. We actually went as far as hiring a director of change management in our organization, because again, I said this earlier, there's nothing worse than a million dollar piece of technology that's not being leveraged. It's nothing but an overgrown coaster. From my perspective, our change management team, our marketing team, work very closely with IT and innovation app development to make sure that we understand what's coming down the pipeline. We understand how it's going to impact our reaps, but we actually even present at every office meeting to say, Hey, here's coming up in it. Here's coming up in technology and here's how it may affect you. And one of the big things that we tend to do is we create champion groups for each technology platform. What that does is it allows, not so much me, the technology person, everyone no wants to hear from the IT guy.
(55:15):
What you do want to hear from your coworker, a tax supervisor or tax partner, or an audit supervisor on a partner that says, Hey, I've been using this thing that the tech guys may, it's amazing. It cut down, it did this screen. In no time, word of mouth from someone that does what you do is goes beyond any marketing message that I could possibly put out, and that's our plan of attack. From that level, from the top down, it's all about making sure that we understand that sable change. Having an ROI calculator and having a KPI calculator really shows the value of everything that we're investing and it shows us where we made bad investments as well as great investments so that this way we can double down the great ones and avoid the bad ones.
Jon Herstein (56:02):
I love the example of the showcasing people who are taking advantage of these technologies so that it's not, the message is not coming from you as a technology leader, but from someone who's actually driving work improvements in the business. We actually do something very similar at Box. We have a lunch every single Friday for the entire company and every week we showcase something that someone's done. We call it box on box, but taking advantage of box AI to make their work process a bit more efficient or a bit better or higher quality. And those have been really, really resonating with the company because again, it's not our CIO telling everyone how great it is, it's actually a salesperson or a CSM or a consultant saying how great it is and it's a big, big, big difference. So let me end with a fun one then. When some of these things go really, really well. How do you celebrate wins for an AI project or digital project? Across the employee base,
Amel Edmond (56:58):
We celebrate what wins on a monthly basis. We have a great team that actually highlights all of our great innovators for our team, and this is by team, I mean globally throughout the rhythm ecosystem. If anybody comes out with a great idea, number one, there is a spike in giving for that great idea. And if that great idea goes to production, there's another great stipend for it. So again, you get everything from gamification to excitement. We do a publication and that person is celebrated and at the end of the year, the most innovative person literally gets to go on stage with our CEO and is celebrated by the entire fur and get a giant plaque, actually a giant crystal statue saying that they are the innovator of the year. Actually, there's no better bragging right to say, listen, I did this. I impacted the fur. And creating that sense of pride and that swell within everyone, especially someone coming in out of college thing, you don't have to be a partner to get this. You can be a staff one and to have this staff one that've actually won this award. It is a tremendous honor to be part of that, and it just creates that sense of inclusion at our firm, and that's one of the things that I was always proud of. I joined with that sense of inclusion.
Jon Herstein (58:16):
That is a very, very cool tradition. So please continue that. Maintain that. We definitely find that showcasing these wins and giving credit to the people who are actually doing the work and coming up with these ideas is really the way to go. So congrats on that. We are short of time and so I just want to say Amel, thank you so much for all of your insights for sharing the time with us and some of the advice that you provided to the folks who are listening. We're all on this journey together. We're all learning together, we're making mistakes, we're learning as we go, and hopefully improving and your insights are invaluable for that. So thank you very much.
Amel Edmond (58:53):
Thank you for Adam. Appreciate it.
Jon Herstein (58:56):
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.