Julian: Hey everyone. Thankyou so much for joining the Behind Company Lines podcast. Today we have DanMallin, CEO and founder of Lucy. Lucy is a one-stop AI powered knowledgeplatform focused on helping organizations find the answers they're looking forthrough all the data and organizations owns and licenses.

Dan, it's so exciting to chat with you.As we were chatting a little bit before the show, not only am I excited aboutlearning from your experience as you've seen kind of technology evolve over theyears. You an early employee at Apple. And, and I'm sure that that wholeecosystem was completely different from what we're dealing with now, especiallyas AI and all the tools and technologies that companies can use, eveninternally to really improve processes, gather information, and overall be moreefficient and effective.

Obviously, we'll get into some reallyfun stuff loosely in what you're doing with ai, but before we get into that,Dan, what were you doing before you started the.  

Dan: Well, I, there's athing about founders and, and I, I am truly a serial entrepreneur, so, I, I hadthe opportunity back nothing like dating myself, but back in the eighties towork for entrepreneurial ventures such as Apple and Microsoft.

But at a, but at a very different time.Right. Today we see these as, as global BMOs and, and, and again, we, we were Iremember at Microsoft, we all fit in a room. Yeah. You say it that way, right.So, so, been doing. Many things along the way. Yeah. And and actually was inthe midst of selling my last company and with again, my current partner.

So there's a few of us, and and so wewere selling it. Leaving one of the partners behind. And and we were a bigSalesforce company called Magnet 360. So, one of the leading Salesforce implementers.And so we were just brainstorming, saying, Hey, what is it that what is it thatwe should be doing next?

And, and we ended up, Coming up with theidea of this, and this is six plus years ago, AI is just coming out. Yeah. IBMWatson beating, Keith Jennings and, and the, the, the Jeopardy challenge. Yeah.Just mind blowing, uh uh, in the day. So we're like, okay, how would we applythat technology?

Yeah. And again, we came outta fortune1000, big business, uh uh, this world. And so there is so much data Yeah. Thatexists inside companies that is so poorly leveraged. Yeah. That, that we justthought, Hey, that's, that seems like a great, great opportunity. Yeah. And andso we've embarked upon the process of really building out this awesome tool setthat, that that just, what if I could ask questions of all the data in mycompany and just get answers.

Yeah. And that's where we're.  

Julian: It's incredible tothink about even. Obviously we jump into Lucy and AI and go into that topic,but in particular with the experience of, the early startups that you've been apart of, how has maybe startup philosophy around scaling, about customeracquisition, has that changed at all or has that pretty much kind of built intothe playbook of how to grow and scale and adopt your technology across,customer base and, and your total kind of addressable market?

How has that change?  

Dan: Well, it depends howyou're funding Sure. Is, is really part of the answer. Yeah. So, I mean, andactually in, in Minnesota where I am, we created something 19 years ago calledthe Minnesota Cup. And so this, it's a entrepreneurial new new ideas,breakthrough ideas, competition.

We've helped 20,000 entrepreneursthrough this process. And I've read more business plans than most VCs. We'llsay it that, say it that way. But the reality here is the best thing you can dois not take other people's money. And I'm not saying that's always an option,and I currently have other people's money, so I don't wanna, I don't wanna makeit sound.

But if you can do that, you can operateyour business differently and you can do, and you're empowered to do differentthings. Yeah. This, the growth model in the VC world specifically is apply,apply these dollars we're giving you, invest them all quickly and as, grow asfast as you can. Mm-hmm. That said, they're looking for home runs out of, outof companies and they're as likely to drive you to success.

Or bankruptcy, right? Because both ofthose outcomes are their finish. Right? And and so, and, and again, VCs are ahuge part of the world. I I have their money, so I don't wanna, I'm not herebashing them. Sure. I'm just saying it affects how you run your business. Yeah.And and you have to be careful though, because you don't wanna run outta moneyor you don't wanna and I had some great great mentors.

Along the way and and board chairman andother investors who are like, Hey build a good business. Right. And the restwill come. Yeah. And so you gotta make sure that you, if you're investing ingrowth, You have to make sure that you have a knob or dial that you can turnback. Right.

And make yourselves just a profitableenterprise at any way, at any level along the way. Yeah. And that's a goodbusiness. And so if you're, if you're losing money on each unit Yeah. You can'tinvest in growth and make that work. Yeah. If you're making money on each unit,but you're investing in getting more people to buy them, then you.

Make that work. Yeah. And, and it all,and, and it works out. So, the other piece is deliver value to your customers.Do it with excellence, customer service and those things haven't changed,right? We, we have to be about just doing the best job ever. And, and it's notthe technology stack that wins, right?

It's the customer success, customerservice stack that wins on top of awesome technology.  

Julian: Yeah. And how has thatbecome, obviously now kind of shifting more towards ai, how has that becomemore of a reality for companies who were strapped for, resources? You can'thire a bunch of customer service agents because they're costly.

And as a founder you can only take somany customer calls depending on the volume of people who are using, what haslike the, the, the evolution of AI now recently been able to, in your mind, kindof optimize for the, the processes that most businesses have as just like anormal function.  

Dan: Well, certainly we'veall I mean, we've all experienced the AI chat bot, right?

Where you log on somebody and they'reanswering questions and they're, they It's a funny use of the word AI becausethey're really programmed outcomes. Right? Right. So they're, they have athousand things they know how to answer. Somebody manage those thousand in aprocess. And then the chatbot itself is using natural language recognition.

So therefore it's ai. Right. In general,it's, it's, it's a very low usage of ai. Sure. And, and obviously in the, inthe last 60 days, everybody has learned about chat, G P T, which is actually AIin a language model. Right. That is answering questions from the corpus ofknowledge that's in Wikipedia and some other internet sources.

And so, but, but for companies, it can,it's not about the internet. Yeah. And that's where everybody's kind of lost.Is it really? I don't want answers in the internet and Google could have gottenme yesterday's answers and chat. B GTP is getting me tomorrow's answers fromthe internet. I need answers from the data that I.

Already invested in creating, that's ourinternal esoteric data that is our, IP and and answers to questions that aren'ton the internet. And I don't really want the Internet's answer. I need ouranswer.  

Julian: Yeah. And, and what aparticular obviously, want you to describe that technology a little bit more ofa loose in, in how it's integrated and kind of what it allows businesses to do.

So we'll give you a chance to give yousome context there, but, How was, how was information gathered previously forthese large companies and, and how much was being lost in the inability toaccess that information?  

Dan: So here's some amazingdata points that Yeah. That are just mind blowing. So the first answer is,anybody who's been at a company knows that.

PowerPoints or presentations are createdat a volume that you can measure in thousands or tens of thousands per day.Wow. Right? Yeah. So at a Fortune 1000 company, they may be making 5,000 newPowerPoints every day. Yeah. So, and that's just to set the scale that thenthat says, I wrote a power.

Presentation six months ago, and I canbarely find it on my laptop. Ha. Yeah, so, so that's just a little case ofpoint, but here's some data. Within 30 days of a PowerPoint being created in anorganization, it is stored for seven plus years. And never opened again, 90%likely, right? 30 days, 90% are never opened again.

Within 90 days, it's over. 98% are neveropened again. So what's the act of, what's the act of putting together aPowerPoint? It takes a knowledge worker, mm-hmm. To research and take the datathat exists in the company, convert it to knowledge in the form that'stransferable through graphs and charts and, and content.

Mm-hmm. And then it is presentednormally once. And never used again.  

Julian: And why, why, why isthat? It, it, some of us would think, oh, maybe the, maybe the knowledge isn'trelevant.  

Dan: But is that the case?Well, the, yeah, the, the knowledge isn't accessible. Yeah. It's in aPowerPoint. It might be stored on a, on your hard drive, it might be stored onyour shared server.

It might be shared on SharePoint orOneDrive or Box or Dropbox or wherever you Yeah. Your company storesinformation and, and actually somebody who's come into the company they, they.They have the situation, which is okay, everything you need to know is on theSharePoint. Go find it. Right. And anybody who's been told that you, you can'treally find anything.

Yeah. And know what it is. And then theother thing is if you try to use, Classic search. What happens is you get alist of documents and it might be a hundred or a thousand documents, right? Butlet's just say you only want, you only are gonna look at the top 10, and let'sjust say that there's only only 10 pages each.

Which isn't, that's not the averagenumber of pages. Pages are 50, kinda number, but, so 10 documents. 10 pages,that's a hundred pages of reading to get your answer. Yeah. And and then wealso know that people in the enterprise are searching for information 25% ofthe time. Really. So if you're a knowledge worker, you're actually spending 20to 25% of your time looking for stuff.

Wow. And worse yet, 80% of the time youbother other people to help you find the answer. So not only am I wasting 20%of the time I'm in interacting with, another 5, 10, 20 people. Yeah. Askingthem to help me find the content that I'm looking for. And if you're lucky, oneof them says, oh, I did something on that and here's the PowerPoint I did.

Cause that's what. What we knew, and wedon't think of it as kind of an iceberg, right? The knowledge that you knowexists is the top of the iceberg. The knowledge that's in the company is thebottom. 90% of the iceberg that you don't, you're unaware of. Yeah. It mightnot have been done by somebody I know.

It might not have been done by somebodyin my work group or business unit or division, or, And so there's just no wayof exposing the content that's out there.

Julian: Yeah. And what goesinto training a sophisticated model? Obviously gathering information from textover the internet is a little bit more feasible, versus going intoapplications, reading through documents, actually ingesting the information,evaluating it, and then, outputting whatever you're looking for.

What goes into the, the sophisticationof building something that integrates with all these different softwares,platforms, and everything that's involved in a company's e.  

Dan: So the, so the, again,the, there's a level of where, where is this stuff stored? Yeah. And so, weattack it with the concept of what is the use case for this kind of set of, ofknowledge that we want.

We identify the. Sources that now. So itmight be on these three SharePoints. Sure. It might be in some purchased orlicensed data from, Mintel and, and not, business intelligence insider orwhatever. And then we, I identify the user group. So even as you're thinkingabout rolling this out, it's not like a big bang.

We don't just connect to all the dataand Yeah. Connect to all the users and, and cuz it's, it's really not the, theway to, to drive this. It's, it's one use case, one set of data or sets ofdata. Yeah. And then the users associated with it and you just keep, cyclingthat use case u data users use case data users, and all of a sudden kind of anenterprise, an enterprise system, but the, the way it works is really, reallymind blowing.

So you have 10 SharePoint sites. Weconnect to those 10 SharePoint sites. We just monitor the ads, changes anddeletes inside there. And so, the, initially all those documents are read,watched, or listened to. Yeah. By Lucy in our case, and she tags. Indexes andcreates metadata. Everything inside there.

Yeah. But we don't do it at the documentlevel. We're doing it. And go back to my PowerPoint example. We do each slide.Yeah. So if you have a hundred slide deck, it's tagged a hundred differentways. Yeah. Each page. So we can bring back the slide, not the deck. Wow. Andthat's how you get, but but that can't be done without ai.

Yeah, it's, IM. To do that. And, andmost systems, most are you gotta upload that content in it? No, we just connectto it and monitor it. Yeah. Most systems you have to up, you have to tag ityourself. And here are the 10 keywords or 20 keywords I added to this deck.Yeah. We're doing, 50 keywords, 50 concepts, things.

Per slide. Yeah. And so we get to a, a,we can actually find that needle in the haystack a across huge, huge data sets.And and so, and, and again, I don't know it exists. I don't know that where it,where it exists. I don't know how to log into Mintel even though we have alicense. I didn't know that we had a license.

Yeah. And and so, so again, it it's theubiquitous kind of access to all of the data sources. Yeah. At one point. AndI, and one single question. And then the other thing we do is we also exposedata things or systems. So if you have a, data in power, bi Tableau, thosekinds of tools. Yeah. Or esoteric third party tools or internal tools thatyou've built.

You can ask again, a single naturallanguage question. Lucy then queries all of those repositories. Yeah. All ofthose third party data sets and all of those systems. And she does, and shebrings back just the dashboard in the system. But what she does is give you thetop 10 answers from across a thousand different places.

Yeah. But, but again, going back tothis, has anybody heard. G p t recently, and Lucy actually has been had G P Tin the product Generative ai. Yeah. For over a year. And what we do is we takeall those answers, feed it into G P T, and she writes a synopsis of the answer.Wow. And so, so we give you, here's the synopsis of the answer.

Here are the 10 sightings of where wecame. And then just to blow your mind one level further, because most peopleare disturbing other people to find information. And do you know how theydisturb people? It's the message. They might use Slack or Teams. So you canjust at mention Lucy in your Slack conversation and she'll answer with thatsynopsis in.

Julian: It's incredible tothink about, the, the way all the information can be gathered. First of all,thinking about, when a document's updated and tracking, that information'scompletely different than having kind of a read protocol, which is amazing tosee that kind of feature built into it, but also how much training or educationis needed to be done in and prompting ai.

And prompting to, to actually gather theright information. Do you go, I do find that that's a huge kind of learningcurve for a lot of these internal, employees and these organizations toactually prompt it to find the right thing.

Dan: So the prompt is, itis really, it's really interesting.

We, we, we have been working hard onnatural language. Yeah. So the the examples of what people. Ask and, and key,they think in keywords kinda searching historically. And then and then I talkedearlier about customer success is such a big part of this. Mm-hmm. What happensis you ask que Lucia question, the first thing she does gives you the answers.

But the second thing she does is did youfind what you were looking for? Yeah. And yes or no. Right? Right. And so ifit's a no, it tees up you. We say, Hey, tell me more about what you werelooking for. Right. And that tees up a workflow around ensuring that The datais there, the da Lucy's doing the right things.

Lucy understands the acronyms. So it,it's part of the learning module. Yeah. It's also, but here's something thatstarted happening really quickly. We realize that somebody might ask a questionlike energy drinks. Yeah. And then we say, did you find what you're lookingfor? And they say, I was looking for millennials, uptakes of energy drinks inthe uk.

Yeah. Okay. So what we started, and thatwould've. Teed up a workflow on our side, right. To make sure that that'sanswering. But what we started realizing is people describing what they werelooking for actually asked the question they should have asked in the firstplace. Yeah. Right. So, so we're actually a, as part of that, we apply thatanswer or that secondary question to see if it gives you what you want and wetee up the workflow and, and make sure that, that you get there.

But the truth is, It's, no, tell me moreabout what you want. And then you give a, an an explanation. And this is a casewhere, I don't wanna say more words are better cuz too many words are aren'taren't better. We wanna, we wanna focus on Hey, ask a really good question.Just like if I was asking you, Julian.

Yeah. I wouldn't walk up to you sayenergy drinks. Because you'd look at me go, what? Yeah. What do you want? Doyou want an energy drink? No. I wanna know about millennials usage of energy,drinks or uptake over the last five years as compared to fruit drinks. Again,the better the question you asked, the better the AI comes up with an answer.

Yeah. And the idea in the AI side is,hey, it's natural language, so we wanna understand the. But the other side iswe've got all the content that we've already indexed. Yeah. And it's also anatural language understanding. And so it's a big matchup of all of that to beable to get you that. And again, we do that against all of those systems andthen we write a natural language synopsis.

Yeah. That you can just read and, andand should be able to give you that answer.

Julian: Yeah. What, what'sbeen the challenge in terms of building trust within, say, a largeorganization? Maybe afraid of privacy, maybe, doesn't think the technology cangather from the re from the information that they have available.

How is that kind of, I guess, pushback,how have you combated that, that pushback of trust and, and actual ability ofthe technology and, and what are ways that you're able to communicate the valueto companies that, that are considering your te.  

Dan: Yeah, there's so muchto, so first of all, security and, and information security and trust are justpart of the every conversation.

Right? Right, right. And and we're ISO27 1 certified, which is a a rigorous process to ensure that we follow greatsecurity procedures. Right. And and, and, and we do external. Testing andpenetration testing, trying to break into, we hire people to try to break intoour system, literally as, as part of, as part of that process.

But so, so again, we're, we're, we havebest practices. The clients are very aware and have. Teams that certify us intheir worlds Yeah. To do that. But one of the things we don't do is thedocuments that we are referencing never leave your servers. Yeah. So if it's inSharePoint, it doesn't get loaded into Lucy.

It's still in on your server unmodifiedand completely available. So we never take possession. We're not the system ofrecord. We don't have anything to worry about there. The second thing, Which isreally important and it re relates to, again, some of these technologies, youdon't have access to everything on SharePoint at your company, right?

You have access to this SharePoint, notthat SharePoint. Yeah. You have access to this five SharePoints, not these fiveSharePoints. So if you and I both have different sets of access and ask thesame question, Lucy actually gives the answer only from the documents that youare allowed to know about.

That's That's incredible. Right. So,and, and, and if you think about it, of course, that has to be the case, right?Because, my vice president of my division has other knowledge that I shouldn'thave. Yeah. And has access to documents about the reorganization has that themerger has a, things that are happening that at those levels that happen insidecompanies.

Yeah. And so, and my manager has accessto different information than I do. And so, and again, and then there's HRkinda content that might everybody have access to. So we, we, we. Availablecontent that's appropriate and available might be in my geography, but not yourgeography. Yeah. Again, hr, well the laws are different by state, by country andso all of which has to apply to these models.

Yeah. And, and again, that's, that's apiece of of mind blowing. And then the last piece, which Lucy does is rememberwe always tell you where the answer came from? Yeah. You know when, when youget an a synopsis answer, but it says, Hey, we, we got this answer out of thisPowerPoint, page 36 of this PowerPoint and out and minute 13 of this video.

And and the, and the third pdf page 12of this pdf. And so you can. That this is real real data. Yeah. From realsources that you trust. But the other thing is because we're not using theinternet, and, and I'm not sure if you're aware, but you can ask the internet aany question and get validating content on any side of any argument.

Yeah. Right. So, so the fact that. ChatGTP lies to people. Isn't that hard to believe because the content is comingfrom this vast unknown repository. Right. But our content comes from yourinternal repository. Yeah. And we give you the references so that. Think of itas the footnotes. Yeah.

In your research paper, we actually giveyou those. So it's, yes. You should make sure that they valid and they'reunderstand. Yeah, the content's coming and Oh, Julian wrote this paper, so I, Ithink it's valid. Yeah. But the other piece is I might also wanna ask Julian,and so you can see who wrote the document.

You can also get access to the, and wealso help you find the experts Yeah. Inside the company too. So I'm looking fora specific type of expert. One way is they wrote the document. The other way iswe actually help you. Inside kind of the HR system too.  

Julian: Yeah, it, it'sincredible to think about, the internal, like a function of an enterpriseorganization and I, I think you proposed a question before the show o of thechange of enterprise work and the standard of enterprise work.

Being, significantly changing. Well,what was the standard before and, and what in particular has been changing inthis working ecosystem. Obviously, I think a lot of organizations just ingeneral, structurally are changing to hybrid and remote and offers differentchallenges. But outside of that, what standards are being changed and what werethey before?

Dan: Well, again, you hadaccess to. Some sub subset. There's, there's data even on onboarding newemployees, right? In, in your first week, you're told maybe 50% of what youneed to know. Yeah. And, and your retention rate is very low. Yeah. Right?Because you're overwhelmed. You're overwhelmed. And, and then again, the restis, figure it out.

Ask the people around. Well now I'mworking from home. There's nobody around me. Yeah. I can't run into somebodygo, oh, I was asked, I had a question about this. Cause there isn't a run intosomebody. Yeah. Kind of, kind of model. Now we've replaced that with the Slackand teams and other things, and that's where we, but we, we tend to do that to20 people at a time.

Right. Yeah. Tell me about this. And andso you, the right person may not have been around to answer, they might be toobusy to. And, and, and again, you're, you're either taking up other times or weteach people to ignore, questions and slacker teams because they're, they'retoo, there's too much going on.

The other thing is the speed of businessYeah. Has turned up right. And and actually the speed of technology. We, wetalk about, dog years and now we're in like, I dunno, is it gerbil years? Idon't know what,

Julian: it's almost insectyears at this point.

Dan: Right? I mean, it'sthe idea of, oh my, it's just happening so fast.

That, how could I keep up with it?Right? In any You know the, it's not like we published the procedural operatingmanual and things and it lasts for five or 10 years. Right, right, right. Wehave to adapt to all of these cycles that are happening at a, at just adifferent. Pace. Yeah. And so if you don't have a system that that does that,and so the idea of being, Hey, I changed that thing, I store it or updated onthe server and then, as, as soon as a couple hours, it's, it's live andavailable.

Yeah. And, and really it, it can befaster it's just mind blowing and globally available. Instantly in multiplelanguages and all of the kinds of things. Yeah. That never happened in thepast. We used to wait annually for the trade show to announce a product.

Right. Right. That doesn't happenanymore. We, we don't wait for the trade show. Yeah. When's the, let's get itout there. Let's, let's cycle it. Even the concept of SaaS software versuson-prem software, SaaS software, we, can be updated. I, it doesn't get, butcould be daily or weekly, right?

I mean, that's, it's usually, monthly orquarterly. But, but the on-prem, we would do, after three years we go, oh, weprobably have to get that update. Right. Right. And, and then there's a processto it and, and everybody, and let's just say it, a thousand customers using it,that means a thousand IT teams had to install it.

Yeah. Versus in SaaS, it's just one teaminstalls it, manages it and keeps it.  

Julian: Yeah. And shifting alittle bit more towards, the business and the outcomes, what's particularlybeen exciting for, the last few years, kind of recent growth, how manycustomers do you have? What's particularly excited about who you've been ableto service thus far?

But what, what are you particularlyexcited about now that AI is, is really popularized? People are way moreinclined to adopt it because it's new, it's innovative. What's been excitingabout the traction you've had to this point, and what are you particularlyexcited about in terms of the next kind of milestone to.

Dan: Yeah, I mean, we we'rein a really cool situation. So we are an enterprise software package. We workfor the Fortune 1000. And I, I have to be, somewhat cautious. But one of ourclients as an example who talks about it all the time is PepsiCo. Yeah. And,and PepsiCo, we're an enterprise platform inside PepsiCo they call Lucy Adathere, ADA.

After the eight 18th century ADALovelace, who's the first programmer, and, and and so. But the, but all of ourclients, you would recognize their logos. Yeah. We'll, we'll, we'll say it,we'll say it that way. And and so watching the, the Expanse and even over thetimeframe, going from hundreds of users to thousands of users to tens ofthousands of users, right?

And and, and truly being an enterprisesystem where you. And we have companies that have a hundred percent of their,their organization on or other, or divisions or different places. But we're inI mean we're, we're I think in 32 countries today where people are using Lucy,where we have, again, this, this enterprise scale.

And, and the opportunities. Stillendless. Yeah. Even inside the companies that are doing things. Cause I talkabout again, new use cases. New things and, and how does that apply? Well, itcould be marketing and analytics data. It could be HR data. Yeah, it could be.But it could be customer success.

It could be call center and operations.I mean, just think about a global manufacturer and they have plants and. 30countries. Yeah. And each plant has operating manuals and things for theirmachinery and. And, and it's in the local language. So in Italian, I need, Ineed a, I need somebody to be able to ask questions in Italian, in Spanish, inFrench and Portuguese in, and, and across the globe, right?

Yeah. And right now they might have aroom inside the manufacturing plant that has all the manuals. And sosomething's I need to maintain something something's broken. And some guy'sjust walking into the room looking at the manuals until he finds the. And thengoes back out and tries it and, and it's an iterative process.

And again, if you could just ask Lucyhow to do something and she brings forward here, are these in three differentmanuals? Here are the pages. Take a look at 'em. Figure it out. Go do it. Yeah.You take hours into minutes. You can take days into minutes. Yeah. And, andagain, I talked about HR and digital learning and onboarding and salesenablement.

So the use cases just continue to expandare just continue to expand. Yeah. And somebody asked me a little bit about anentrepreneurial. We can't necessarily attack all the markets that Lucy is validin, right? Right. We're an entrepreneurial organization. We gotta spend ourmoney, we gotta do the right things.

Right. A across those things. But ourclients keep a adding more to it, like, could it work here? And then we do alittle test with them and it, and, and without doubt it just, it, the answer isyes. Lucy doesn't have limitations. We just have, the, the entrepreneurialventure staffing and, and, go to.

Kinds of thoughts that we have to makesure that we're smartly spending our money and, and attacking, broad, widemarkets.  

Julian: Yeah. Yeah. If you thinkabout whether it's external or internal, what are some of the biggest risksthat Lucy faces today?  

Dan: Well, it's funny, we,I, we talked about the, the speed of of of the innovations and changes andcertainly chat gtp no technology has ever moved as fast as chat gtp.

As I said, we were already working withgtp for two years. It was in our product for a year. It was, we were last chatour G P T three 3.0 sure. 3.5 turbo and then, and now four. And and so that'sall, that's all part of again, how fast we move. Our, it's really interestingcuz the conversations around GTP have really opened our market for us becausemore people realized, They really want to and need to get there, and they don'tknow how to get there.

Yeah. But the second component, which isintriguing, which is it organizations are going, wait a minute, we could dothis now. Yeah. And, and the short answer is, Can and should is a, those aretwo different questions, so they can do some of it, but, but they reallyhaven't thought about the scale and scope of what they want to do.

Yeah. Because, Microsoft's making theseofferings available. But they only do some of what we're doing. They're onlyinside Microsoft infrastructure and, and, and not out, not this u unifiedversion of that. And then the other piece, Here's a little secret about your ITorganization.

Where's the customer success portion?Yeah. Yeah. Because because whatever they build, you need to, to support and weneed to, what happens when a connection goes down? What happens when we wannaadd new data? What happens? How do we make sure that we're getting all theright data into the system? And that's really where the internal side alsobreaks down, right?

Yeah. So, So is it the best use? I'llsay it differently. I know my IT organization could. A word processor. Yeah.But we buy Microsoft office. Right, right, right. I don't, I don't need towrite stuff. I don't, I could write a payroll system. I could write an e r Psystem, I could write all of these things.

And the reason we choose to buy them isbecause the cost of ownership is substantially less. Yeah. And, and, and again,in SaaS software, just installing it. Right, right. Installing and managing andupdating and all those things. Well, we centralize that across all of theclients and. There's a little fud caused by all this conversation fear,uncertainty, and doubt.

Yeah. Of where they want to be, butquickly they realize, no, we actually do need to buy it. Yeah. We actually needto buy it from a company that's doing it at scale. Right. We wanna, we do wantcompanies that listen to us and, and certain we have a customer advisory board.We, we wanna hear what our customers want and need.

Yeah. And, and, Our clients have createdwhat the product is together with us. Yeah, right. We've modified and made itbetter every step of the way. And and we continue to work with our clients tomake Lucy the most powerful answer engine. Anywhere. Yeah. And and, and again,they bring in new data sources, new connections, and we just make sure thatthose become part of the platform.

They bring in new content types and newopportunities. And then, and Lucy continues to learn. Every time somebody asksa question, everybody time somebody gives feedback. Yeah. And every time and,and. Lucy's continues to get better and will continue to get better with ourclients.  

Julian: Yeah. If everythinggoes well, what's the long-term vision?

Dan: Total worlddomination. Again, we're entrepreneurs. Right, right. Our job is to, toinnovate and create great technology. I told you I'm a serial entrepreneur. Itold you we have investors. All of those things mean we're gonna sell. Yeah.Yeah. And I should say it could mean go public or other things, but I don't, Ithink that As true entrepreneurs, you have to understand that going public isnot off.

It shouldn't be the goal always. Mm-hmm.And and it actually trips up lots of companies because, it costs millions ofdollars a year to be public. Yeah. And if you're a hundred million dollarcompany the overhead is too high. Right. You if you're a billion ormulti-billion dollar company, that may be an opportunity.

Sure. But even then, Finding the righthome inside the right organization for our product and products is always gonnabe the outcome. And certainly that's what our investors expect. They don't wanta dividend check, they want a liquidity event check.  

Julian: Yeah. Yeah. I alwayslike this next section, I call it my founder faq.

So I'm gonna hit you with some rapidfire questions and we'll see where we get. I always like to open it up withthis question. What's particularly hard about your job?

Dan: I, I, so first of all,I, I don't, I think I have the greatest job in the world. I get to do what I doevery day. I have just the most awesome employees. I get to mentor and lead andlead them. We get to. Again, innovate for the world's best companies and and weget to affect their speed of doing business.

I get answers in seconds, not days orweeks or months, and and so I help them make better decisions faster. And so I,I don't, I, I don't have a hard. To my job.

Julian: Yeah, I love that. Ilove that. I, I was, just thinking about, obviously there's always concernaround AI and, and especially because of the velocity and what it's able to do,there's always like concerns about jobs, but I, I'm more, I'm more excitedabout the possibilities it's gonna allow people in terms of a tool to be ableto get answers more quickly or, What are some ways that AI is not being usedyet that you're particularly interested in?

Maybe other companies tackling it or, ormaybe even you might tackle it in the near future. Is there any, is there anyway it's being underutilized that you'd like to see the technology be, be usedfor?  

Dan: Well, so first of all,we keep using the words AI or letters ai. Sure, right? Sure. So I wanna defineAI as I see it, versus the world sees it.

Please do so. Everybody's thinkingartificial intelligence. And artificial intelligence is part of what I think itis. But I actually, we use the words augmented intelligence, and so we're aboutmaking knowledge workers better, faster, all kinds of things. And again, thoseapplications are infinite as, as it's just as what we're describing, but it'snot about operating.

A manufacturing piece of equipmentmonitoring things that are happening. Yeah. And automatically adjusting things.And that's a good use of ai. I'm not saying don't do that. Yeah. That's notwhat we do. We are focused on, hey, you have 10,000 or a hundred thousandworkers. We want to really, really help them do their jobs better.

You have a million customers. We wannahelp them have a better experience working with your company. Yeah. And. Andso, and, and there is a level of, it applies to every, every job, everywhere ifwe're doing it right. Yeah. And yet companies, as I said, and we're workingwith them to pick the first use cases.

Yeah. Pick the first data sets and pickthe first users. And and again, I don't, the Big Bang doesn't actually work aswell here. Sure. Because the AI is a learning system, and so you're better offstarting. A thousand people and then getting to the hundred thousand people.And you can do it quickly, but you still need that first thousand that goesthrough the process and gets there.

Yeah. A and, but companies, companiesare not there. Yeah. In most companies they, they do not and have not deployedthings. They're working with legacy technologies. And, and sets. And so, andthen this is where chat GTP has been awesome for the overall concepts ofartificial intelligence because everybody now realizes what.

This technology can do, and actually I,I'm gonna say this out loud, open AI built chat GTP not to be a consumer toolfor 10 billion users. Mm-hmm. They built it as a proof of concept of what isthe art of the possible. Yeah. And so now people have been exposed to the artof the possible, and it has real world use cases right now that it's cool andit can work.

But again, as I said, we're focused on awhole different data set. We're ho focused on different use cases, we'refocused on all kinds of things, but people now understand and are demanding of.It organizations of their organizations, the capability to leverage the datathat we have. Yeah. To find answers in seconds and to be able to bring forward,I mean, what used to happen is a executive would come down, they'd ask a questionof a group of people.

They'd say, we don't know the answer.We'll get back to you. They come back a week or a month later with the answerand the executive says, I already had to make that decision. Because theydidn't have a week or a month or longer to do a data collection to do all Ineed the answer now. Cause I'm making this decision today.

Yeah. I have to make this decisiontoday. And so that's where this power is just totally gaping. Yeah. Or or, or Iguess this newly exposed power. I said, I've been doing this for years. Butmost people are just learning about what it can do now.  

Julian: Yeah. Yeah. I alwayslike to ask this question cause I love how founders extract knowledge fromanything that they ingest, whether it was early in your career or now, whatbooks or people have influenced you the most?

Dan: So, I, I, I mean, Iwanna say I've read all kinds of books. I, I it's interesting. So from anentrepreneurial perspective, actually I'm gonna reach back. I can just grab it.There's a book called Traction. Yeah. And it's written by a gentleman namedGino Wickman. And this is an entrepreneurial operating system.

They call it eos. Yeah, yeah, yeah. Andand so the whole idea of EOS is this is how to run the business, run meetings,track your things. It has nothing to do with product market fit, it's all kindsof things. But it's an awesome book. And this is how we operate our companiesand our all. In in past companies too.

But the cool part is, when you get anoffer from one of my companies, you get the book. Yeah. And we send it to youand we say, Hey, read the book cuz this is how it's gonna work. And your weeklymeeting is described in here. And our quarterly of planning sessions aredescribed in here. The annual planning session is described.

And so, and it's not that it's the bestoperating system, I don't care. It's a fantastic operating system and there's abook, and so I don't have the manager create it or, or, or teach that. Yeah.And so, and, and again, when I say that I could reach into my credenza andthere's 20 of them here because I hand them out like water, and I and I makesure that everybody knows you need to operate your company in a predictableway.

And so this, and, and for employees,they actually appreciate it, right? Here's what it's it tells you how we judge,manage, make sure we have the right people in the right seats. Yeah. For acrossyour organization. And right people is defined as the people that follow andlive our core values.

And right seats is, do they get it? Dothey want it, and do they have the capacity to do it? Right. And it, I mean,again, it's, it's a, it's a bible in how to operate an entrepreneurial ventureand it doesn't really stop there. But you know, because bigger companies canuse or leverage the concepts there, but they, this is about  

Julian: entrepreneurialism.

Yeah. I love that. And I always like to,I know we're coming to the end of the show here, and I always like to ask thisquestion to make sure we didn't leave anything on the table. Is there anythingI didn't ask you that I should have or that you would've liked to? Anythingleft on the table here?  

Dan: Well, There's thething about entrepreneurs and and I don't, I dunno if you'd asked this, butagain Minnesota Cup, we've helped 20,000 plus entrepreneurs.

I I've interacted with all and youtalked, then you ask questions about books. Here's my belief. Entrepreneurs areborn, not created. Yeah. Books. And knowledge and education and experience makethem better. Yeah. Right. And so, and I go back to Michael Jordan, he was anawesome basketball player, right?

Coaches and training and repetitive madehim to the level that he was, right? And so you gotta put in the hard work, yougotta be willing to do the, all of the work necessary. We all can learn, a treethat stops growing dies. So the second I think I know everything, I'm going todie. Right.

And and so the idea here is so I'm notsaying don't read a book or don't, you can get better, but there is a, we havea, we have a society today who thinks everybody's an entrepreneur. And I, and,and I hope everybody can be and they can do things, but there's a thing aboutcreating. Real companies, real products, versus, and I'm not saying aninfluencer isn't an entrepreneur, but they're not running a business yet.

They're just doing something and, and totake something to scale and do all kinds of things. There's something in yoursoul that says, I want to get up every morning work, 60, 80, a hundred hours aweek. Yeah. I want to risk everything. Including financially. Emotionally, myrelationships and all.

And then I get to, work until I'mfalling asleep at the keyboard and do it again tomorrow. Yeah, right. And sothere you could say it's a, it's a form of insanity or stupidity, but it's athing. And and nothing drives entrepreneurs like, The opportunities to succeed.And, and the best entrepreneurs are doing it for a mission, their mission.

And, and so enabling corporate knowledgeand, and, and being the, the ultimate answer engine for every company in theFortune 1000. That's what drives me every day. I do it for our clients. Andthen the other piece is, and, and for the employees. Yeah. And again, I'm a, I,I got gray hair and season entrepreneur, but nothing makes me and my partnersmore.

Excited than watching our formeremployees in their own entrepreneurial ventures. Yeah. Succeed. Yeah. And, and,and hearing about other companies and and, and their liquidity events or theirsuccesses. And so, you, you gotta cheer on your own. Right. Right. And so we,we help and mentor and grow people.

We help inside, inside our company. Wedo it for others outside our company. Yeah. And, and again, in Minnesota, we'vedone it for 20,000 people and that. That's the payback.  

Julian: Yeah. It's incredible.Obviously to, to learn about your experience and, and where you've come fromand, and how you kind of got to, running loosely, but also what you've learnedalong the way.

I can't wait to share this with theaudience for so many different reasons, not only philosophically, but withstrategy and, and how to approach things. But last little bit is where can wefind you and be a supporter of you as, as a, as a founder and, and. Go out andseek out information from an individual like yourself.

What are your LinkedIns, what are yourwebsites? Where can we find you and be a supporter? And also inquire about waysto, to build better.  

Dan: Yeah, for sure. Well,I'm, I'm Dan Mallin on LinkedIn. I'm and Lucy is Lucy ai. And just check, checkout what Lucy's doing. There's some videos there, and of course I want you totell everybody you know, that that if you're looking for real.

Augmented intelligence in theenterprise. Lucy is the answer.  

Julian: Amazing. Dan, it'sbeen such a pleasure having you on this show. I hope you enjoyed yourself andthank you again for being on Behind Company Lines today.  

Dan: Thanks Julian. It'sgreat to meet you and great to share

‍