#161 Natural Language Data Analytics
Subscribe to get the latest
on Thu Sep 21 2023 00:00:00 GMT-0700 (Pacific Daylight Time)
In the latest episode Darren Pulsipher sits down with Steve Wasick, the CEO and founder of InfoSentience, to discuss the power and potential of natural language data analytics. Steve, who comes from an unconventional background as an English major turned screenwriter turned lawyer turned tech founder, brings a unique perspective to the field.
#ai #generativeai #embracingdigital #edt161 #challengesinnaturallanguageprocessing #fantasysportsapp #naturallanguagegeneration #highqualitycontent #technicalbackground #outoftheboxthinking #pushingboundaries #diversity #crossdomaincollaboration #innovativeideas #infoscentience'ssolution #dataanalytics #naturallanguageaisystem #conceptualautomata #datasets #revolutionizing #businessesanalyzeinformation #futureofdataanalysis #naturallanguagereporting #flexibility #tailored #differentindustries #customized #specificcontext #jargon #dataanalysis #revolutionizingindustries #sportsanalytics #stevewasick'sjourney #innovativeapproach #entrepreneurs #techfounders #unconventionalpaths #successfulinnovations #embracingvariability #context #poweroflanguage
Steve recalls his early project—an app for fantasy sports that aimed to provide users with not just statistics, but also the context and stories behind the numbers. This led him to the field of natural language generation, where he faced challenges in acquiring and delivering high-quality content. Despite not having a technical background, Steve’s diverse experiences allowed him to approach these challenges with creativity and out-of-the-box thinking.
Darren praises Steve for pushing boundaries and bringing a fresh perspective to the field. This highlights the importance of diversity and cross-domain collaboration in generating innovative ideas and solutions. Steve’s journey serves as an inspiration for aspiring entrepreneurs and tech founders, proving that unconventional paths can lead to successful innovations.
The conversation also delves into the capabilities of InfoSentience’s natural language AI system. Steve explains that their technology breaks down events and stories into their constituent parts, providing a better understanding of complex concepts and their relationships. This analytical engine, based on conceptual automata, allows for the synthesis of diverse and complex data sets, revolutionizing the way businesses analyze information.
Furthermore, Steve emphasizes the flexibility of their AI system, which can be tailored to different industries and customized to meet the unique needs of each client. By understanding the specific context and jargon of the data being analyzed, Info Sentience ensures that their AI system provides accurate and relevant insights.
In conclusion, the podcast episode highlights the potential of natural language data analytics in revolutionizing industries such as sports analytics. Steve Wasick’s journey and innovative approach serve as an inspiration for entrepreneurs and tech founders, reminding us that unconventional paths can lead to successful innovations. The future of data analysis lies in embracing variability, context, and the power of language.
Hello, this is Darren
Pulsipher chief solution,architect of public sector at Intel.
And welcome to Embracing
Digital Transformation,where we investigate effective change,leveragingpeople process and technology.
On today's episode,
Natural Language Data Analyticswith special guest
Steve Wasick, CEOand founder of Info Sentience.
Steve, welcome to the show.
Darren Thanks so much for having me.
Hey, we had a fascinating conversation.
When was it? Three or four weeks ago.
You show me some of the tech.
You guys are doing it.
Very fascinating stuff.
I got to hear a little bitabout your background.
I think it's an interesting,fascinating background.
I want you to tell our audiencea little bit where you came fromand why you're doing what you're doing.
Yeah, So I'm a tech founder andgenerally,
I think to have takenthe technical lead within my company,
But I have sort of a weird backgroundfor that because I started outlike I was an English major.
I was actually at the University of Hawaiiand I was trying to be a screenwriterand was in Hollywood for a few years.
I ended up going to law schooland was in law schoolat Northwestern in Chicago,and in my third year I had this sort ofcompletely random idea for fantasy sports,for a product, for fantasy sportsthat just became sortof my overwhelming passion.
And so I worked on this this appbasicallywhen I was finishing up law school.
And after I graduated, I raised some moneyand started this companyand ended up having to sort ofin some ways learn how to be a programmer.
I took programing in high schooland really liked itand kind of kept up on it,but I certainly wasn't any goodand so definitely wasa bit of a rough, rough road to do that.
But I think that and we'll talk more aboutmy company later, but I think that thethe space that I was in in terms of,you know, innovating was new enoughthat there wasn't really any benefitto kind of having a technical background,at least withinwithin the innovation space,because it was all brand newand and in some ways having an Englishbackground and having a legal background,which is a lot about logical rulesand relationships and things like that,
I think really helped me out.
So, so yeah, that's,that's kind of a different path
I think for afor a technical founder, at least.
But you know, I'm still hereso was not that.
Well you'reyou're every computer scientistworst nightmare
I'll just tell you that right upfront.
Yeah. I'm a software engineer.
Oh, okay. Yeah.
And you guys driver's crazy becauseyou got all these crazy ideas, and,you know,you could figure out how to do it.
And as computer science is reallygone now, you can't do that.
But you do it, right? Yeah.
And then you drag people along andthey go, All right, we can fix this up.
I love when people do cross domain stuffbecause you bring new ideas.
You have that diversitythat brings it into the field.
And when I saw your stuff, I could tell.
I could tell there was a different mindbehind it, right?
Meaning it was more and we'llget to that later, but it was moreuser friendly, which is a big problemfor a lot of technical people.
They don'tthey don't do things user friendlysays, Well,you got to learn how to use the tool.
I think it was the difficultfor my first employeewho was trained, you know, incomputer science because I had ideas.
And looking back on them, a lot of timesthey really were the correct idea, right?
Like they were coming from this placeof having struggled within the fieldthat I was working with,which is natural language generationand kind of understandingwhere the wrong turns were.
And so I was like, Hey,we got to do it this way.
But it was also that packagewas also coming along with like,
I don't know what a dictionary is,so I'm just using arraysand some sort of weird,you know,
So it's like hard for him to be like,how could this possiblybe the right solution?
Because the syntax is so stupid,you know, like that.
I it's almost likeyou get this challenge of somebodyif somebody is like talking in a secondlanguage, you know,and it's like they can be brilliant,but they're having trouble with it's it'shard to say, Yeah,with the actual communication.
So it's kind of like that.
So I think thatthat was some struggle struggles for uswhen we were starting off of himjust kind of being like,
I don't think this is right, you know.
And so that was definitelya welcome feature.
Yeah, you'reyou're a softwareengineers worst nightmarebecause you push us outsideof our boundaries.
And we all get stuck in our boxes.
And I love how you pushedoutside of the box on, on, on this.
When you showed me the product,
I was like, Oh, that's pretty clever.
Thank, that's. That's really clever.
Pushing way, way beyondwhat us computer scientists think about.
So let's talk about the promise spacefor First off,you said fantasy sports.
If you're in Chicago,it's got to be baseball.
It wasactually I'm sorry, it was football.
Yeah, because I was.
I was just in a fantasy sportsleague in law school.
And, you know, I was really busyand I was just kind of like, man,
I don't even know what's going onwith my own team, let alone, like,what's going on in the league.
And I was like, Boy, I'd be really coolif there was a way to kind of like providecoverage like, you know, unique,interesting information about my league.
Just like you follow real sports,you know, like where you have picturesand headlines and here's the top storiesand what's going on and insightand bring it inlike the really cool context of like, wow,you know, like this team got blown outby 30 points in the playoffs last yearand now they blew them out by even more.
You know, in this game, you know,and it's a big revenge gameor something like that.
That's what makes sports actuallyinteresting, right, is all the contextand the stories behind the stats,not just necessarily the game itself.
I mean, you know this from watching.
If you're watching a sports gamewhere you don't know the teamsand you have no rooting interestor any context, it's like it's usuallypretty boring, you know, unless you'rejust an absolute fanatic for a sport.
If you just drop in, you know,to a game, it's not that interesting.
So so that was my idea.
And I started working on itwhile I was still in school.
And as I startedsolving a lot of the challengesthat that came alongwith actually getting really good contentbecause I started offjust kind of doing like a Madlib and like,okay, here's who won and here's who lost.
And I was like, Oh, let's try to addsome more interesting pieces.
It was really achallenge to start thinking like, Oh,you have to deal with sort of repetition.
We could talk about maybe more that later,but there's a lot of difficultiesin creating a really in-depthlong piece of textthat's completely fluid in termsof like what it can coverand how it can fit togetherand everything else.
And as I started solving those challenges,
I started realizing, Hey,actually this is more than just likea fantasy sports thing.
This is this is a general sortof analytical engine and synthesis engine.
And and so that started meon the path of being like, okay,
I actually start a company around thisbecause I thought
I was really compelled by the idea.
So so let let me reiterate,
I think I heard you sayyou took a lot of statisticsand lots of informationabout about the league,about your team first.
And youand you wrote itin something that you could easily read.
Instead of looking at graphs and thinkbecause people, even though we think wewe get a lot of context from pictures,we get some context from pictures.
It's the written word or the or someonespeaking to you where you reallyhear the emotion.
You hearall all the things involved in it.
So that's what you didwas you wrote somethingthat was easy to read,not normal computer generated.
Like numbers and yeah, player with number.
No, no one's going to read that, right.
Yeah, I thought that was pretty ingenious.
I think you know, it's to yourto your point about even graphs chartsyou still have to tell the storyin your head, right.
Like you have to see a lineand it's going downand then there's something in your headthat has to say that's going down, right?
Like,you have to translate that. Right. Andand that's the beauty of a storyis that you don't write.
Like, if, if, if,if when we talk about good writing,we're really talking about somethingthat allows usto understand the informationas opposed to analyze the information.
When you're reading somethingthat's really well-written,it's it's just like it's giving youeverything you needto actually think about,about what's happening, right.
And not have to come up with what'shappening and even charts and graphs.
And certainly if you have like a row ofnumbers in a in a in a table or something,you got to do the whole thing.
Okay, what's going on here?
How does that fit in to what's happenedin the past or what's happening?
You know, I'mlooking at the West Coast information.
What'show does this fit into like the East Coastor our overseas stuffor whatever it is that you're looking at?
If it's just numbers,you got to tell the whole storyand then you have to think about it,right?
Whereas if it's written, you could justsay, Here's everything you need to know,and then you as a decision makercan then, you know, run with that.
So so that's that's really interesting.
I love how you said well-written, providesthe whole story and color all in one.
When you said that,
I thought of some poetrythat needs a lot of analysisthat we had to do in high school.
Don't get me down that road.
That be kind of funny, though.
Could you imagineif if your analysis spit out haikus.
We could do that.
Hey, work for hire.
So if somebody wants theirtheir business intelligencein haiku format,you know, we'll give it a shot.
Or, you know, or iambicpentameter write all of the same sonnets
Would Tricky match. They wrote.
That'd be pretty funny.
So. All right.
So the written word in conjunctionwith graphsbecause you guys have it in your stuff.
So you're doing the analysis. Yes.
So we all know I needs a lot of datain order to understand other data.
So how are you training youryour analytics engineand all you've done it for sports?
Is it a general analytics engine that
I can run against anything now and then?
How do I have context, all that stuff.
So yeah, that's a good two part questionbecause they both relate to each other.
So we don't really train itin the sense that it's not like an alarm,it's not probabilistic.
We okay, so it's not it'snot a large language model like
GPT or Bard or any of those.
This is a different AI Yes.
Yes. We use what we callconceptual automata,which is all about breaking apartan event or storyor interesting factinto its constituent parts.
So I use an examplefrom from sports to start.
Like if you say, all right,we know what a team is, right?
You have this idea of what a team isand you can add all these sortof characteristics to it.
And we can say that it's sort of similarto an organizationor it's made up of players.
You can look at all these,the subcomponents, everything else,but then you can alsohave an idea of a team winning a gameand what that means, right?
And it might mean different thingsfor different sports, right?
But you can havea general concept of a teamwinning a game and then you can also havea general concept of a street, right?
Which is just a series of eventsthat happen over time.
And then if you put those all together,you can have this idea of a teamon a winning streak, right?
So you put them all togetherand now in our system that when thatwhen all those things come together,everything else in the systemcan relate to it because it understandsits subcomponents, right.
So it understandsall the things that it's made upbecause as you get moreand more complicated,you can't have like a bunch ofif end statements or templatesor anything else, it'sgoing to fall apart instantly.
So you have to be able to have have,you know, the intelligence embeddedon these subcomponentsand give it the intelligenceto be able to combine with other things.
So even you could say like, all right,they were on a winning streakand then they lost this game.
Well,that's a broken winning streak, right?
So it's likeand then that can fit into a narrative.
But, you know, you can'tyou know, you can't say that, right?
You can't say, hey,you know, we're on a winning.
You know, they were on a winning streak,then they losttalk about that game and then say, hey,you know, they actually won last game.
And that was the third straight gamethat they'd won. Right.
Even thoughthose are two different things.
Hey, they won three straight gamesand that they you know, as of last week.
Or they they broke a winningstreak of three games this week.
Like those are two separate storiesin the systemif you were to build them outindividually.
So and that's just a small example,but there's a things like thatwhere it's like you have subcomponentsof stories, you have the storiesthat are very clearly relatedto each other or inor even in opposition of each other, suchthat likeif this this sentence followssomething that's negative, right?
And now it's positive,you might have to have a little transitionlike that said, you know, they are,you know, doing great in this other stateor something else like thator, you know, that's thatsaid they do have their leading scorercoming back, you know,so that might change it, right?
Like everything relates to each otherand human beings.
This is you know, our superpower is beingable to just have all these conceptual.
Yeah. To understand the relation to.
We just slotted right in. Right.
But that's something that traditionallycomputers have struggled with.
And so our system, you know,can now handle that in ways that thatthat allows like really in-depth,complicated narratives to come togetherin ways that are not templated.
We are not like a modelthat's company, like we are an AI company.
And the good news is that, you know, thatmakes it really flexible and powerful.
The bad news is we can't just apply itto everything right out of the box, okay?
Like if you're talking about sportsand you'retalking about the stock market,it's like our system definitely does both.
Like we have a stock stockproduct, commodity product,marketing stuff, healthcare stuff.
We are all the verticals basically.
And wecan't just sort of jumpinto a brand new data set, a brand newanalytical question essentially,and have it ready to go.
However, there's a good chancethat you care about thingslike something being on a streakor something being on a trend orsomething,you know, moving a big amount that way.
And then the other waythose concepts are already in there.
So the concepts translate,but the context. Yes.
Is different for the different. Yes. Yes.
So that's that's where you needyou have professional servicescome in to your to your companyand help, you know,establish the context for,for the data that you're working on.
How does that work?
Typically,we just. Work with our clients, right?
And usually, you know, a lot of timeswhen we're working with somebodythere already is writing on the topic,right?
Like sometimesthey're doing manual reportsand so you can just look to see like,okay, what are they care about?
You know,what is the jargon that they use?
What is you know, what are the types ofthings that they talk about togetheror that,you know, tend to step on each other?
All these different things, you canyou can get a pretty good senseor something like sports foryou know, I'm a sports fan.
Some of my employees are.
So we already know that.
Same thing with stocks Like you havea pretty good idea with that.
But there are some use cases for surewhere we just need to talk to the clientand just basically interview them aboutwhat is it that you care about,
Do you have any examples?
And then we take that and then use that toto create the content.
So there is somethis is really when I heard you talking,
I said, wait a
I is a lot of a lot of English majors.
I have a lot of friendsthat teach at university and Englisheyes kind of scaring them a little bit.
And even even writers are havinga hard time.
And the strikethe strike in
Hollywood with and I mean, that's dearto your heart a little bit, right?
Yeah, yeah, yeah. Working on that.
But what, what I heard you said was yourunderstanding of the English language and,and your training as a lawyerhelps you understand language really welland how how we understand language.
And that's the foundationof what you've developed here is,you know, how how concepts relateto each other in language.
It's I think it's brilliant, right?
Because you found a new branch of where
English majors and and writers can go,and that is to help eyes become even morecognizant of the way that we think,which I think is brilliant.
Another question around this
How flexible isthe tool that I can I can tune it myselfor do I need you guys to help metune it for my my special case?
Because I can tell youfrom my experience talking to customers,they all think they're special.
They all think that no one's ever doneanything like we're doing.
But that 20% or maybe even 5%,they want something special.
So how easy is it for mein my company to do?
Or do I need to sit downwith one of your guys to reallyconfigure the model or configure the airto do what I want to do?
Yeah, we don't have an enterprise version,so you definitely have to talk to usand I will.
If you're out there,
I will definitely make you feel specialbecause I would say all of our clientsreally are their own unique uniformswhen it comes to their data.
Like no matter what, likeeven if it's a similar data type of thingthat we've had in the past,there's always quirks.
And I think that this is actuallyone of the fundamental problemsof a data analysis that's out there rightnow, is that it's really hardto have a sort of one sizefits all structure for any databecause once you do that, thenyou kind of cut offa lot of the unique parts of,you know, like you, because it's like,let's say you have a company,
I should literally just talkingto somebodyyet last week about this exact issuewhere they have a bunch of differentmanufacturing plants. Right.
And they have a bunch of different sortof data structuresin their manufacturing plants, right.
Where it's like,oh, sometimes they refer to it as an air.
And other people,they have a sort of broken down.
And other timesit's like, well, is this really an air?
It sort of it's like it had to goback on the line, but then it got fixed.
So that's kind of a somewhat you know,it's likethere's all these different thingsand it's like if youif you're trying to say,okay, let's just synthesize,let's just have a standard data set.
Well, now all these manufacturing plants,which are all a little bit differentand doing different things,they all have to put their informationinto the standardized set,which is difficult enough in one moment,but then they're constantlygoing to be changingand adding thingsand switching things out.
And now they're either going tojust not be able to add that data, right?
Or they're going to you know, they're justthey're going to try to wedge itin or something.
It's like you need a system thatunderstands that that things are flexible.
Right? And so that'swhat kind of our system does.
And that's why we have to sort ofdo it on our own a little bit is becauseit takes a lot of training essentially,to understand how to, like,pull out the commonalities,how to put things into an ontologyessentially,and thenalso how to get that into our systemso that can do the analysis.
So it's so you can.
Just actually embrace you, you embracethe variability in the data, right?
Then you can normalize thatwith context, which I think iswhich is really difficult to do.
And I love the story of the manufacturerbecause you're right,as long as there's humans involvedand not just humans, but also machinesthat come from different vendors,the data is going to be different.
I mean, what is an errorin one manufacturing plantis very differentthan in another manufacturing plant.
But at the same time, if you're runningall the manufactured manufacturing plants,you want to be able to have a reportthat maybe tells you like,yeah, errors are up, right?
But like, what does that mean?
You know,you might be able to have something where,yeah, all these are like a type of errorand so we can have a numberthat says here's an errorwhile at the same time, if that managerwants to look at the informationthe way they want to look at it. Right.
Which is like, well, we have like redosand we have like partial errorsor whatever it is, they can still seethe breakdown as they see it, Right?
So it's like because we can do thisas human beings,like if we're as human being and we'relooking at the way that they set up,you know, this one manufacturing set plantsets up their data.
It's no problem for us to understand,
Oh, these are all types of errors, right?
Like it's easy, you know,
And if some other manufacturing plant,they have a few different things,that's easy for us as human beings.
So it's definitely possible.
And that's what we that's sort ofwhat's driven us.
It's like, well,this is definitely solvablebecause human beings go aroundand do it all the time and we don't like,you know, have our brains explode.
Does not compute or like this thing isthis one was capitalizedand this one wasn't.
So now it's all broke.
It's like,that's not how human beings work.
And so it'sdefinitely possible to do it that way.
And that'swhat we've we've endeavored to do.
So pretty cool.
So the verticals starting in sports,you moved to financials,now you're movinginto manufacturing as well.
What are the what are the key verticalsthat you guys have everapproached or attacked?
Yeah, basically, you know, sports,we do have a medical product,just just one right now.
We have a couple of differentfinancial products and we've done retailstuff, marketing stufflike like analyzing marketing campaigns,really, you know,our system can be applied to anything.
But from a practical standpoint,we really need somethingthat's pretty complex in termsof what somebody needs to read, right?
So I always use the example ofif if you're talking about,all right, I'm from I'm in Indiana.
If the Colts are playing the Titansand all you care aboutis who won,then you don't need our system, right?
Because you could just see it was 3221.
Yeah, that's that.
If you want to know what happened.
Well, that's a huge, complicated questionbecause there were thousands of thingsthat happened in the gameand there's literally millions of waysyou can contextualize all those.
And so if you're trying to say, like,give me five paragraphs to know whathappened, you need a really sophisticatedsynthesis of the data.
And that's a lot of times businesses,you know, maybe they have a lot of data,but really they only care about sort oflike three metricsand then they're good to go.
So that's they don't need us.
So if you have somethingthat's complicatedand then also have somethingthat's a pretty big scale because it doesbecause we are modelingthis stuff on our own,because we do have to sort of set upa data ingestionplan, a distribution plan,like these things take time.
And so it really only makes sensefor us to do itif if it's going to bea pretty wide scale.
But a lot of timesbig companies have that right.
So it's not just, hey,we want our sales report every quarter,but actually we want everybodyon our sales team to be getting a reportevery week detailing what's happeningand how it fits into the largersales figuresand all these different things.
If you need something like that,then that's a situation where,you know, we could be very helpful.
So let's talk about the engagement modela little bit.
So if I engage with you guys, your teaminterviewsus, It's it's an engagement, right?
How long does it takefrom the time that I say,
All right, Steve,we want to do this at Intelto look at all of our manufacturing plantsand all that stuff, right?
And so what's that look like?
How long does that take two to typically,
I know it's going to be,you know, for every customer,but are we talking a year engagement orengagement, three months.
Probably 3 to 6. Okay.
From start to finish. Yeah.
And you guys get right downdeep in our data.
Talk to us. Right.
It's a collaborative thing, right?
Yeah, but I it's pretty rarethat anybody has to spend that much timebecause usually it's
I mean, I would say 2 to 3 hours tops,you know, to get us startedand then maybe, you know,another few hours as we iterate,you know, But it's not like it's not likewe have to sit down and because it's likeusually it's pretty straightforwardin terms of what we're looking at.
Like we have a sense of like in general,okay, you care about this going up or downor these subcomponents, right?
And then it's and so it doesn'tusually take that long to sort offigure out thepeculiarities of your particular needs.
And in that so I don't want to give peoplethe impression that they're goingto be like having to sort of like train usand whatever systems they usually just.
Have The experts, you guys understandit already.
Yeah, and usually it's pretty clearmore or less what they're interested in.
And so it's it's more aboutjust understanding the subtleties, right?
Even some of the jargon,the language understanding, you know, justhow long they went through itfor who they want to targeted towards,
I mean, things like that.
But it's not like a whole,you know, multiday sessionor something to get us to know.
What to do.
Also, you mentionedreports, right?
Because, I mean, that's the ultimate goalis to understand your datain a textual and graphical formbecause you have you know, they're bothhow how often can I see can Iget that information in real timeor are they generated reports?
I mean, what is it? You know what I mean.
Yeah, I mean, scale is not a problemfor us in any way, right?
Like whether that's number of reportsor timing.
For instance, you know, the fantasyproduct that we do with CBS Sports,you know, that's regularly producingmore than a million articles a week,people that are going out to folks.
And a lot of those are even producedin a short period of time.
So we're like we're literally producing,you know, hundreds of thousandsan hour at pointsand we could go faster if we need.
So so I could do real time.
I could do real time information.
I mean,a productthat we're doing with the Chicago
Mercantile Exchange, which is
I think you saw the demo that we have.
Oh, yeah. Yeah, yeah.
But that we're bringing thatthat wide to the public.
So basically we're creating a websitefor the Chicago Mercantile Exchangethat covers every single commodityin their system in real time.
And it's going to be a very different way,
I think, to get information,because peopletypically when you're on a website,it's like you have stories, right,that somebody wrote and maybe thosestories are 3 hours old or 4 hours old.
You know, they come from rightin the financial space.
It's like sometimes those are accurate,sometimes they're not.
And, you know,
I think it's going to be very differentsort of looking at this websitebecause you're going to have this hugeamount of content and videos and chartsand everything on every single topic,except it's just pretty much like nowthis is all the information right now.
And if something changes,the headlines are going to change.
And whenever you log on, that'swhat's going to be there.
And you can even put in your preferencesin terms of like the types of thingsthat you care about, like the commoditiesor the types of metrics even.
And like all of it'sgoing to change for you.
So it's like these are.
Customized for, for me,what I'm looking for.
So that that's pretty cool.
It's not like a static report.
And one thing you showed me,
I remember I thought was really cool.
I have a story in front of me.
It generated based off of my preferencesand then I can highlight,you know, something in the text.
I says, This is really interesting.
I can highlight that.
It'll show it on the graphthat matches it.
Then I can click on itand get another storybased off of that small amount of datathat I said, Yeah, I'm really interestedin this part of the story.
Click and details in story format.
Again, that drill downcapability to me is is brilliant, right?
Because I can now traverse my datareally in story form,which isthe analysis is happening on the fly.
To methis this is really about the interface
I know I know what would cool technologybehind it's making it available.
Yeah but when we talk about the real winon this to meis that you guys are interfacingwith a human insteadof a human learninghow to interface with the computer.
Very, very. The opposite nicely.
I think we're moving in that direction.
A lot of other A.I.,which to me is fascinating.
Yeah, I agree with you.
I mean, I think that we havehad such a great responseto the demo because of that, even thoughthe technology isn't necessarilythat much better than it was beforewhen it was just the story is because nowpeople can really see intuitively like,
Oh, this is what's happening.
This is the flexibility,this is the power behindit in ways that just saying, oh, here,read this reportthat we did on fantasy sportsand then read this reportthat we did on stocks.
I mean, you know,it's like people are like, oh, cool.
But like when they actually get to, like,choose your own adventure,essentially through your data,it's it's powerful in a waythat that isn't wasn't there before.
I love that.
Choose your own adventurethrough your day.
Yeah that's that's the tagline
I, I think
I think it's pretty brilliant frankly,because we've seen data martsbefore and in some respectsthis is kind of like a data marginwhere I've set the data up so I can easilytraverse it as long as I have a Ph.D.in data science.
As long as I understand the data.
And what you've done isyou've you've delivered the data toprofessionalsthat don't have to understand the data,but maybe understand their their context,whether it's A.C.
Sports or commodity commodity exchange,
Write to me that one.
That one's brilliant, right?
I can look and see.
I get my story.
I'm really interested in oranges,so I'm going to drill down and, hey,the orange market's doing,we are going to drill down in thereand maybe it gets me to some data onforecasting for the winter this winter,whatever the case may be,it's all right at my fingertips,which I think issomething uniquethat you guys bring to the table.
Thank you. Yeah,we think it's pretty cool.
And we think when peopleget to actually do it on their ownthat there'll be a pretty good responsebecause this is going to bebasically public.
You have to have like a CME accountwhich is free to sign up for it.
So yeah, and being able to drill downagain, I'll just caveat a little bitthat like because it's asort of public websitethat's not supported by an individualprocessor, you can't necessarilydrill down anywhere you want to go,but there's going to be a lot of a fairamount of that, likelike all the kind of like typical basisthat somebody be interestedin are going to be coveredand sort of pre-builtso that when they want more informationon a topic,it's like, Oh, we already wrote that.
Here's your web page.
It's written like, So yeah, there's yeah.
Except that it's rewritten every,you know, five, 10 minutes, right?
So it's like every time and againwith various preferencesand all sorts of other things.
So it's like dependingon what you're interested in,it's like that articlesessentially waiting for you.
But then if we,if we had a client that waswanting to have this supportedin terms of like the total free freedomversion, they can get that too,because I mean you sort of generateit actually generate the content.
Literally on the fly then.
So real quick,how do people find out more about it?
Where where do they get information,you know, about your companyand your products?
So my company's name is InfoSec Giants.
I don't think we mentioned that.
So I'm very bad at marketing,but no info centered on superbecause that's not even a very good name.
But whatever it is, but it is so info
Ascension's is is our website.
My name is Steve Rancic.
I'm on LinkedIn so people can followme on LinkedIn. I try to dosome posts aboutnatural languagegeneration and things like that,but but yeah, you can just go in thereand request a demoor reach out to me through LinkedIn.
But yeah, we,we, we still have lots of fun things
I think we're going to be doing.
Yeah. A lot of different verticals. Yeah.
So I think if people I'd loved, you know,every time I give the demo
I think I told you this, you know,and I asked you, I said, Hey,have you ever seen anything like this?
Right? And everybody said, no.
Like, nobody's like, oh, yeah,this is like what so-and-so is doing.
Or like, Oh,no, You know, everybody always says, No,
I haven't seen anything like this.
And we built all the technology upourselves.
We're not using anybody else'stechnology anywhereother than we use a similar line.
Right? But like, that's it. You know,we use a similar, but that's about it.
So everything else and this run.
This can run on prem and in the cloud,not run like a run anywhere, right.
And you can, you can run it anywhereand we deliver anywhereand the data againbecause we have to kind of model it right.
Like it doesn't matterwhat structure your data is like.
We never tell our clients like,
Oh hey, like, you got to put these thingsin order, and these things areit's just like, just give us access.
And then we hit itand then we put it into our ontology.
So you can do structuredunstructured data. It doesn't matter.
That it has to be structured. Yeah.
So that's whyit has a structured structure.
But it's like you can have ten differentsilos and again it's like, Oh,this manufacturing plant, they do thisand this one they use that andthat has some complexity of the project.
But, but we deal with that againbecause we want like the reasonthat it's typically in a different formator in a different structureis because it's actually different, right?
Like sometimes, yeah,it's like the field name is the same orit's like just a field name.
But most of the time the reason that likethe structure of the data is differenthere than here is because there's actuallysomething different going on.
See, that's a very different approachin most data scientists use.
Most data scientists come in and say,
I need to clean and curate all my data.
It all needs to be normalized.
And what you're sayingis embrace the variability, understandwhy it's there, and and use it.
Yes, I think is brilliant.
So, Steve, this has been fascinatingto learn more about you and yourand your company.
Thanks for coming on the show, Darren.
Thanks so much for having me.
This isn't this is a fun talk.
I think that's a take.
Thank you for listeningto Embracing Digital Transformation today.
If you enjoyed our podcast,give it five stars on your favoritepodcasting site or YouTube channel,you can find out more informationabout embracing digital transformationand embracingdigital.org Until nexttime, go out and do something wonderful.