Consulting Rainmakers

This AI made our calls longer and that was the win

Brendan Daly · Daly Ventures ·

Brendan Daly is the Founder and Managing Director of Daly Ventures. This Consulting Rainmakers episode, hosted by Mikulas Zach, is about one idea, in Brendan's words: This AI made our calls longer and that was the win.

In this conversation, Brendan explains why buying the technology is the easy part of AI adoption, how an AI meant to shorten calls ended up making them longer and opening up sales, and how he picks which AI use cases to build first.

Watch on YouTubeSpotifyApple Podcasts

About Brendan Daly

Brendan Daly is an engineer and AI adoption advisor, a former Director of Technology at National General Insurance (Allstate) and a three-time UiPath MVP. He set up Daly Ventures in Ireland eleven years ago out of necessity: after helping found DayBlink Consulting in Washington DC, he was waiting on a green card and could not work for the firm he had helped build, so he started his own consulting practice. Repeat work for insurance clients led to a full-time role setting up their AI and automation program, where he scaled real-time AI coaching for contact center agents from 30 people to 2,500 within six months. Today Daly Ventures advises leaders on AI strategy, adoption and automation, and runs a small venture studio supporting early-stage startups.

Expect to learn why change management decides whether AI gets deployed enterprise-wide, why giving people access to the tools and permission to experiment matters, how AI and speech to text replaced random spot-checks of ten calls with a ranked view of every call, how a one or two day workshop turns scattered ideas into a prioritized backlog, why the best AI ideas often come from the people closest to customers, how a few quick wins justify bigger AI investments, and much more…

Brendan Daly: LinkedIn · Website

Chapters

  1. 0:00Cold open
  2. 1:02Welcome
  3. 1:20Starting Daly Ventures Out Of Necessity
  4. 2:41Why He Has No Elevator Pitch
  5. 5:48Where AI Adoption Really Starts
  6. 8:22The Biggest Surprise Of Running A Firm
  7. 10:04The AI That Made Calls Longer
  8. 15:14Choosing The Highest ROI Use Case
  9. 21:45Quick Wins Before Big Investments
  10. 23:20What Is Next For Daly Ventures
  11. 24:38Where To Find Brendan Daly

Transcript

Machine transcript of the conversation between Brendan Daly and Mikulas Zach, lightly cleaned. Spelling of names and terms may be off.

Mikulas Zach

So joining me today is Brendan Daly, the founder and managing director at Daly Ventures. Brendan, thank you for coming on and sharing your insights today.

Brendan Daly

Mick, yeah, great to be here. Looking forward to the conversation.

Mikulas Zach

Me too, me too. So you founded Daly Ventures what eleven years ago? what was kind of like the pivotal moment that made you go on your own?

Brendan Daly

Yeah, yeah, good question. So at the time it was actually out of necessity. So I had been involved in setting up a another management consulting firm in Washington DC that was called DayBlink Consulting. but at the time I actually had I was waiting for a green card to process and had an 18 month gap. So I couldn't work with a company that I had helped founded in the US. So as a result, I had to figure something else out. And at the time ended up setting my own firm in Ireland, a management consulting firm, and then saw some work. My initial project was actually with a company in Canada doing a kind of a technology rollout assessment for them. So yeah, it w it was out of a s a necessity originally. and I kind of just ba I had some experience setting up a a consulting firm before and then kind of did something similar myself.

Mikulas Zach

Got it. And right now, to my understanding, what you do really is kind of like the AI adoption, governed AI adoption and strategy. how do you usually describe it if you'd have you know you'd be in a elevator, what's kind of like your elevator pitch?

Brendan Daly

yeah, no, good question. so I would say like I wouldn't necessarily have an elevator pitch for like in my pocket that I'm just pitching everyone. I think it is there is a lot of opportunity with AI and I think it's very much depending on the context of who your client is. So whether you're in insurance or if you're in you know, transportation or co operations. Like there's so many different uses of it that I think, you know, I would actually first again just like ask people what are they most excited about, what are they most concerned about. and, you know, kind of go from there. But Yeah, I wouldn't say I have any elevator pitch. And again, to be honest, I do think it kind of almost sells itself in terms of just like how fast everything is moving and the capabilities of it that it's not as if you kind of need to convince people that A is a thing anymore. So yeah, I I would say it's again stuh but for me it's more about listening. What are the problems? You know, what are they trying to do? Are they aware of what's currently available? And then then coming back maybe with that elevator pitch of like, you know, how I can help, but yeah, I wouldn't have any kind of predefined elevator pitch as it relates to that.

Mikulas Zach

Got it, got it. And you serve mainly what, insurance and healthcare? is kind of like the the main

Brendan Daly

Yeah, it kind of by default, even though I have done media, I have done hospitality. so I think that's one of the benefits of management consulting is that you do kind of have multiple clients and it can be different industries. so I did get a lot of diversity kind of in the earlier years, and then yes, what happened was I delivered some work for insurance.

Mikulas Zach

Hm.

Brendan Daly

Clients, then repeat work, they brought me back, and then ultimately actually they hired me on full time to help set up the AI and automation program. So yeah, I think it like as a result of me working with them in the past, I kind of got pulled back into insurance and then likewise healthcare. I had done some consulting work in the space, and then after I had left, the one of my former colleagues, you know, had set up a new healthcare venture and you know got me involved and then there was another venture that spun up after that and then I got pulled into that. So yeah, it kind of just opportunistically kind of fell into those areas. But I think and again one of the advantages of consulting is abroad, you know, lots of different industries and then I got yeah pulled back into healthcare and in insurance for the most part.

Mikulas Zach

Talk to me more about how you actually start when you're trying to do a AI adoption in a company. What's kind of like the the the main strategy?

Brendan Daly

yeah, I would say AI adoption, I think people maybe underestimate there is a large change management component to this. so you know, you can buy the technology, but I I do think like there is to get it deployed enterprise-wide. Safely, there is a relatively large lift. so I think people underestimate that. and there is a large change management component. So it's not just the technology you need to get right, it's also the setting expectations, what are the new processes, and getting people comfortable with that. So yeah, I would say it's kind of You know, for successful AI, I do think like there does need to be kind of a larger conversation around, you know, what are you trying to achieve? discussing change management, how are you upskilling people, how are you going to make them feel comfortable. so yeah, I would say that like in terms of I think that's a key piece to it. And then I would say at this point, because it's changing so fast, just giving people access to the tools and being able to play with it, I think is also very important. So they kind of get a feel for what is possible. But yeah, I would say those are maybe two components to like a successful AI deployment adoption, I think. You do need to have kind of clarity around what you're trying to achieve and get the boy in and have a change management strategy. and you do need to understand that there are some technical considerations and architectural decisions that need to be made. so that's on one side of it, but then on the other, I think it's moving so fast that you just need to give people access, get people playing with it, and then, you know, trying to empower them and help them find ways to innovate and make the business better. so yeah, I think that's probably where I would initially start would be upfront strategy, clarity, alignment while also giving people access, making it fun, giving them permission to kind of experiment and come up with ideas and themselves. And then that kind of helps build that culture and helps with that change management component.

Mikulas Zach

Hmm. What do you now that you're you know working and you have a company for quite some time, what's been kind of like the biggest surprise for you in running it?

Brendan Daly

yeah, I would say like probably underestimating there is so much like effort involved with the accounting piece and financial statements accounting. And it is also difficult, like especially in the consulting side, you're trying to deliver work while you're also trying to sell, you know, work and so just getting like you know, it is cyclical, there can be long sales cycles, so Yeah, it can be difficult. Like it's you know, there's it's amazing when things go well and if you can really get an anchor client or you can line up some work, things are great and you get a lot of freedom and flexibility. But again, one of the I suppose things that I wasn't expecting would be s how complicated some of the reporting is and how time, you know, it takes so much time, some of these things that You know, there is really a trade off, you know, when you do have to decide between, you know, working somewhere else where you can kind of focus really on yourself or, you know, becoming really good at project management or AI or whatever that is, versus if you are trying to set up your own farmer company, there is a lot of other kind of accounting, over ahead, compliance that you also need to do. so yeah, just a kind of a consideration, but I think that was there was more involved there than I was originally expecting.

Mikulas Zach

Hmm. And given that you're kind of on the forefront when it comes to using AI, what's been like some of your best use cases for using AI, whether it's internally or externally?

Brendan Daly

yeah, also I would say maybe some of the earlier use cases I had when I was in insurance. that really was kind of what opened my eyes to it as well, just kind of some of the power of this. but I think just like its ability to scale on demand, like we had started off looking to deploy AI in contact centers in a call center. With the original idea that we thought it could maybe cut back on the the time of calls and help reduce average handling time. because the way we had done it was we had AI that was going to be kind of like an AI real-time live coaching for all of the call center agents. So we started in claims thinking, okay, this is gonna help, you know, bring claims information so we can resolve claims faster and reduce call time. But what actually ended up happening was because we were able to help. The customer more, the call was actually getting longer, but now we were much happier and we were able to identify sales opportunities. So again, it wasn't, and that that was kind of interesting, I suppose, with AI in the sense that we kind of went in thinking we could it would do one thing, but it actually enable a whole other opportunity in area. and then that's also then when we got our biggest win. So when we identified that okay, actually maybe there's opportunities on the sales side. what we were able to do was build custom AI models that knew all of the policies we could sell, had all of the the logic of, okay, we can sell, you know, RV policies in these states or motorcycle in these all of these complexities that you know are really hard to teach, especially and and figure out maybe in real time on a a call with multiple policies. We had AI that was able to deploy to all of these agents and in real time come up with Suggestions based on the exact context of the conversation. So, and again, we started out with I think 30 people and it we scaled it to two and a half thousand people within six months. with custom playbooks in claims, customer service, and sales. So again, and I think the impact of that was projected at over 100 million because we were able to not only identify where we could sell more accurately, but also Also based on what the top performers were doing, we could kind of have the AI position the products and help agents with that. So yeah, we could basically almost kind of replicate some of the best practices of the of the best agents with the help of AI. So I think that was a big one and a really cool kind of at least for me in terms of AI, like how whether it was claims or it's customer service or sales or it had a lot of different applications, a lot of use and it had a big impact and it could scale rapidly once we figured out what could work. so that was one. And then just the second one, we did some kind of fraud detection modeling around sentiment analysis and different kind of call analytics and policy analytics. So again, just the just the power of neural nets to identify things that maybe you know you wouldn't typically be able to do. and then the last thing I'll say is maybe one of the the big changes and the kind of wins we had was We in the the claim center, they were trying to figure out like they were doing training at the contact centers. And the way they did it was they would just pick ten random calls, the person would listen in, and then, you know, depending on what they heard, would give feedback, whether it's good or bad. but you might have a thousand calls a day. So they were just randomly picking these calls with no real kind of context or it was just spot checking. So what we were able to do with AI and speech to text was in real time Transcribe the calls so we had all the data, what was said throughout the entire conversation. And then we were able to do modeling on that, so sentiment analysis or do different analysis, and then triage those calls and provide them each with a score based on certain criteria and then rank them. So instead of randomly picking calls, we could give you, you know, the top ten calls to listen for for whatever thing you might be interested in. So whether it's for a really great performance or complicated call or so. Again, much better granularity, much better insight. and again a hundred percent kind of insight into the conversations and the calls versus randomly choosing like less than one percent. so I think that yeah, that was also like a really big win.

Mikulas Zach

Well, that's I think a fantastic use case. And one of the questions as you were speaking that kind of came to my mind is how do you choose which opportunity to take and deploy AI in in a company? Because there are obviously AI has so many different use cases and so many different avenues where you can apply it, whether it's you know sales, marketing. operations and so on. So do you have any kind of framework around choosing the one that has the highest ROI?

Brendan Daly

Yeah, yeah. yeah. So again I would say ROI is definitely a keyword there and I think that is something that w it should be kind of focused on. Typically what I would have done in the past though would not be necessarily like coming in and just do one use case. Like I would come in maybe meet with whether it's you know, the pay if it's specifically just for marketing or You know, depending on the company, there might be a marketing team that's interested and a sales team and a service team and a claims team. So typically there's a lot of interest across an organization. Typically, what I would kind of do would be kind of an upfront scoping session, like an introduction about AI in the industry, you know, what people are doing in terms of like frontier, you know, both within the industry and maybe not even in their industry. Just like these are some really cool things that are happening within, you know, speech to text or IVR or intelligent process. Again, so this is in your industry and even outside of your industry, what people are doing that's kind of really interesting. Kind of teeing it up depending on the kind of level of sophistication of the organization, give a bit of a background about kind of this this is AI, how it works, the different models. and you know maybe neural net or you know depending on the specific you know use cases we're going after but then I would typically maybe try and do like a one or two day workshop type session where you listen to everyone's different ideas and there's kind of a high level scoping. and that just gets people brainstorming, gets people excited. I like building kind of maybe a backlog. and then you can start doing clustering and seeing because then you know, okay, maybe this piece is very hard. But if you do that, there's eight other automations or AIs that it opens up. So I think, you know, not just jumping in with one use case. I I like kind of maybe taking a small step back, giving a bit of a, you know Lay of the land, the current landscape analysis, kind of more specific industry analysis, the the executive stakeholders, whoever, like you know, what are you hoping to achieve? You know, are you trying to optimize for customer satisfaction, or is this a revenue play or a risk play or a compliance play? So, again, there's different kinds of dimensions to all of this, but typically I get all of the key stakeholders together, make it fun, do some brainstorming, get a lot of ideas out there. Try and build out a backlog and then you can start kind of combining ideas, getting people excited, seeing where the the the real traction is. So I think that's like a kind of an important step that is often skipped. Then maybe once you narrow it down, you can kind of put things in buckets. So, like really hard, great idea, but not for right now. But let's capture that. Okay, great, that's on the backlog. Then there's ones that maybe okay, like that's interesting, but that's not it's not there's no ROI, it's not sufficient and it's kind of just maybe a personal thing. There's no real scale, no real impact, it might not be exciting enough. So that's maybe another bucket. And then you've got the ones that are kind of more interesting. So then I'd like to I'd like to maybe spend a small bit more time there doing a deeper dive. That's when you check, okay, who are the who would be involved here? Do we have a clear process owner? is the is the decisions consistent? is there visibility? Is there high risk here? What are the systems? Are there gonna be so again? This is then where you kind of need to do a bit of a double click, a deeper dive. But that's where you kind of mentioned the ROI business case. You know, what would be the level of effort here? What tools do we need? What system access do we need? What data do we need? What's the level of effort? How many people would use this? Does this scale? Is this for our customers? Is this so all of these type of questions? But again, really back to you know, what does this executive team, how would they define success for this AI initiative? And kind of okay, based on that and based on the interests of the team. you know, trying to distill out, you know, maybe kind of three or five candidates. and then for me, I like, you know, I would try and maybe do two or three ideally. Like I don't like maybe putting it all. I like maybe when you can having maybe two or three that you can go. and ones that are you know, kind of lower hanging fruit, not like too too easy, but you know, some lower hanging fruit where you there's a some level of complexity and we you need to get the date in and that, but it's kind of a quicker win, has some kind of impact, gets some momentum going and gets people excited. So yeah, I would say it's kind of a phased approach. first again I would start with kind of ideation, getting people on the same page, sharing, you know, what we're trying to achieve with this, getting people excited. If people are nervous, maybe you know, addressing that, explaining to people, you know, the opportunities to upscale, all the resources that are available, and kind of maybe just laying out the vision of where we're trying to go and what we're trying to do and why we're trying to do it. I think that kind of will help with buy-in. and then I would say kind of again open it up to the team because you know, a lot of the times the people at the the front. working with the customers on the calls, like they're the ones that have the ideas that can really move the needle and have an impact of the company, not necessarily the people four levels up. So again, th there's definitely use cases there. But I would Again, kind of encourage, you know, giving people that time, maybe some space, giving people access, you know, listening, giving them a a tool or a template that they can submit their ideas, because again, a lot of great ideas come up there, that can even be kind of tweaked and refined and enhanced from there. So giving people that kind of opportunity to submit ideas, taking a step back, seeing where the excitement is, the energy is, kind of clustering it, that's kind of the first analysis I would say. And then From their overlaying, okay, what are the systems, what are the data, how does this scale, what is the impact, what's the ROI? And then like once you do that, and if you can get a couple of kind of wins then early, that can then justify bigger investments. So that's what I've c how I've seen kind of larger AI rollout, is that okay, the the executive team is like, okay, look, we we definitely want to play in here, but you know, it's kind of very new to us, we can't just like blow all this money and have no nothing to show for it. So like typically I've seen success when it's very scale focused, clear success. You get a couple of quick wins, you show that okay, we're on the right track, we can do this, hopefully have savings or increase sales or something that actually, you know, generates some revenue or some savings. And then that can then be kind of used to double down and then scale it to okay, we we've just deployed our first three We have a backdog of, you know, twenty five. Let's p what are the next five and ten or ten we want to kind of take on that can be, you know, also kind of raising in complexity. because the nice thing with a lot of AI and a lot of this stuff is, you know, once you solve the the system access issues or the data issues or, you know, so what you can reuse a lot of the components or a lot of the so, you know, the the the first ones tend to be harder. and then it can get exciting then once pe once people see some quick wins and kind of a roadmap and and ways to kind

Mikulas Zach

So what's on the horizon for you and Daly Ventures? What's kind of like the the main goal going into two thousand and twenty-seven?

Brendan Daly

Yeah, great question. so for us again, I I would say we kinda have two focus. One is definitely advisory, so trying to work with teams and companies that are trying to be kind of maybe more on the cutting edge or using some of this technology to make a big impact or do some exciting work. So yeah, definitely, you know, excited to keep doing that. And then we also are involved, we kind of have a venture studio at early stage ventures. So we do a small bit of kind of early stage investments and supporting early stage startups. So yeah, I would say kind of advisory and ventures and just trying to stay on top of you know all of the developments in AI because it's moving so fast and it's becoming more and more capable, which is opening up more and more use cases and opportunities. So yeah, I think it's just trying to stay at the forefront and keep, you know, working on really cool projects, working with, you know, ambitious, you know, leaders that are trying to put some of this technology to good use.

Mikulas Zach

No, definitely. I mean it's changing so quickly. Like I feel like, you know, each week, each month there is something new that's going on. and yeah, I'm excited for the future. And what's the best place for people to get in touch with you, Brendan?

Brendan Daly

Yeah, great question. I would say so Brendan at dalyventures.com. So for anyone listening here, feel free to reach out. Also have Twitter, Be Daily and LinkedIn, or my website, dalyventures.com. There should be some contact opportunities there. So yeah, feel free to reach out, happy to, you know, give advice or suggestions or share resources. But yeah, no, it's I d I think there is a lot of opportunity. It's it's an exciting time and especially for those that can kind of understand both the business and the technology. I think you know, when when you understand what the tech can do, when you you understand and what the business is trying to do or trying to achieve or where it's trying to go. I think being able to connect those you know is really where it starts getting powerful and exciting.

Mikulas Zach

Yeah, for sure. so thank you for coming on, Brendan. I will leave links down below so people can click through and get in touch. I found you on LinkedIn. I think that's the best place. but yeah, thank you guys for listening and Brendan, thank you for joining.

Brendan Daly

Yeah, thanks Mick. Enjoyed the conversation.