The Andy Show – S1Ep37
NICKY PENNYCOOK [00:00:36] Hello and welcome to today’s Andy Show. You’re joining by me, Nicky Pennycook today for hosting. It’s the 10th of June, and today we’ve got a very special guest from New Mexico. His name is Chris McLaughlin. He’s the Chief Product and Marketing Officer. And I’d like to introduce him to you now.
CHRIS MCLAUGHLIN [00:00:59] Good morning, Nicky. Good afternoon. Thanks for having me.
NICKY PENNYCOOK [00:01:02] Hi, Chris. It is a pleasure to have you. How are you doing?
CHRIS MCLAUGHLIN [00:01:05] Doing very well, thank you.
NICKY PENNYCOOK [00:01:06] Great. So just to begin with, it would be great just to get a little bit of background on yourself and what it is that Nuxeo actually do.
CHRIS MCLAUGHLIN [00:01:15] Okay, fantastic. So first and foremost, foremost, Nuxeo is a content services platform or software company. We focus very much on the enterprise Content Management space, digital asset management space and really helping our customers to get more value out of their content. The information that drives our business. I’ve been with Nuxeo about three years and previous to that worked with other ECM vendors companies like DocumentOn, FileNet, Thunderhead. So I’ve been in the space more than 20 years all over.
NICKY PENNYCOOK [00:01:49] Wow, that’s incredible. And say, I’ve got a few questions for you here. So I’m going to dive straight in and to begin with. How important is content within modern enterprises and on these organisations getting the most value from their content?
CHRIS MCLAUGHLIN [00:02:05] Two short answers. Very and no, I would guess that the longer answer. Right. So first and foremost. And so we really like to kind of say content is everything. And when you think about it and how people do work, right? How they collaborate with each other, how they share information and, you know, that’s as simple as things like Microsoft Word, PowerPoint, Excel, that’s all content. Right. So foundationally, content is how organisations do work. And most estimates would say about 80% of all information exists in the form of content. The problem we have for most organisations is that the term content isn’t very exciting. And what it really boils down to is all of this unstructured information that kind of helps drive how organisations do work, how they deliver value to their customers. So when we look at the importance of content from an organisational standpoint, what we’re really looking at is this is information that companies use to drive their customer experience that can be as simple as how we service customers, how we deliver statements, how we deliver correspondence to customers, or that can be as complex as the assets and visuals and images that drive their Web site and their customer experience there. It also can be the type of information that drives core business processes. So when we deal with banks or insurance companies and you think about how you apply for a loan or how you submit an automobile claim, that’s all content. Right? So we talk about customer experience. We talk about bringing products to market. We talk about driving cost and efficiency. We even talk about compliances. More recently, you guys fed GDPR content sits at the heart of compliance with GDPR. If a customer wants to know what information you have about them, 80% of that information on average is going to be content. So very, very critical for organisations. The core issue with content. Why organisations don’t get as much value as they should out of it. Is one for most organisations content exists in a bunch of different silos and two for a lot of organisations it’s very hard for them to find critical information and particularly content when they need it, provide it to customers, use it to drive their business processes and really use it as the foundation for competitive advantage in their organisation. So there’s the longer answer.
NICKY PENNYCOOK [00:04:35] Thank you Chris. So we’re moving on and the use cases that A.I. is best known for content manager may not be at the top of people’s list. How common is these of A.I. live in Content Management?
NICKY PENNYCOOK [00:04:49] I would say today kind of increasingly so. I think most organisations that we talk to right now are trying to figure out how they can better leverage Artificial Intelligence and particularly Machine Learning to kind of address some of those problems that we talked about earlier. How do we make information easier to find? How do we deliver that information, perhaps more predictably, to the people who need it? So a lot of organisations are wrestling with a number of court challenges around Artificial Intelligence and content, and they’re just now beginning to figure out how to begin to apply these technologies to really kind of get over these hurdles and make that information more accessible to them. As I said, the foundational problem with content is that for most organisations it’s just hard to work with, doesn’t sit nicely in a row or column in a spreadsheet. And what we really need to do is kind of help bring more structure to that content, extract data from that content. So that’s what organisations are. They’re starting to think more holistically about how to pull all this information together and get better data about their content. And they’re beginning just now to realise that Artificial Intelligence and Machine Learning can help solve a problem for them.
NICKY PENNYCOOK [00:06:07] Absolutely intersting enough as we actually having a conversation earlier today with someone about A.I., and this is something I know a little bit about, but also I know it’s becoming quite relevant for people as things are changing at the moment. You probably touched on this a little bit, but. I’m going to change your question a little bit. Why is there a need for A.I. Machine Learning and Content Management? But how how important is that? And especially now that times are changing and we’re in a lockdown period. And there’s maybe not as manyas staff around, how have you seen that change? Is there like a real push in need for A.I.?
CHRIS MCLAUGHLIN [00:06:44] Yeah. And I always want to be careful with Artificial Intelligence because a lot of times we get concerned about our technologies like this beginning to displace jobsm right? Or supplant humans in processes? And the answer is no. A lot of the challenges that we solve with Artificial Intelligence, Machine Learning really are things that organisations don’t do or don’t do well. And it’s also the kind of work that humans really don’t like to do. It’s very detail oriented. It’s by distracting data out of information. So it’s one of those things where we can solve kind of a new prop or an old problem in new ways. And without throwing bodies at it and without creating a bunch of work, that, again, isn’t really well aligned with what you want your knowledge workers and organisation to be doing. So when you think about kind of A.I Content Management core core foundation, when you’re managing documents and when your organisation has a very large number of documents or digital assets, that the real challenge is being able to very efficiently find that information. And, you know, if you think about Google, right? And full text searches and things like that, yes, you can type in a term and get a bunch of results back to you. But it’s not a very efficient way of working. And for a lot of our customers who have literally billions of documents scanned images or digital assets. It’s a very inefficient way of working. So what we really want to do is begin to create more structure around these documents, really focus on providing data values that we associate with these documents that make them much easier to find and more importantly, make that information contextual, right? So when we think about data and content, what we’re really thinking about is making that content more accessible, making it more contextual, making it more we usable in an organisation, whether it’s digital assets or traditional documents. We find a lot of customers spend a lot of time, effort, money, recreating information they already have inside their organisation. Now, very, very simply, in terms of problems that A.I. solves, I usually like to break it down in three ways. One is kind of extract and we’re extracting data and information from documents, from unstructured information. And it really this is to solve that problem of how to enrich more data around my content, to make it more findable, more usable, more contextual in the organisation. But increasingly, as we’re working with customers, we’re also seeing opportunities to automate traditional business processes and apply A.I. to look, for example, at a forum and see whether it’s been completed correctly before I passed a form off to a human knowledge work. Right. So here’s a great way for us to increase the efficiency of knowledge work that we do and eliminate a pretty tedious task for a knowledge worker. The final thing is insight. And we can probably touch on this a little bit more later. But what we’re finding now is organisations beginning to put together pools of content, kind of what we call content lakes, and leverage the content and data that they have collectively to get new insights into their business. So there’s some really cool things that organisations are doing with content and data and Artificial Intelligence. It’s beginning to solve not only traditional business problems, but even give them new advantages in the marketplace.
NICKY PENNYCOOK [00:10:11] Thank you Chris, so for people who might… these in a more traditional platform, or that’s looking at me even into A.I. what are the difference between differences between an A.I. based Content Management system than that a more traditional platform.
CHRIS MCLAUGHLIN [00:10:28] I don’t know if we have enough time today to go through all of that foundationally. Okay, and we kind of look at when we think about Machine Learning and specifically we’re kind of talking about Machine Learning here to begin with. Right. How do we train Artificial Intelligence technologies to do certain tasks? And how do we use existing content and data to make that possible? For newer technologies, you really kind of see two different approaches from an A.I. and Machine Learning standpoint. And they’re complementary approaches. So, one, typically there are a lot of public services out there that you can use for different activities. So companies like Amazon, Google, Microsoft provide these public services for Artificial Intelligence. You can do things like OCR. So converting, you know, scaned images or images of text into actual text and data, you can do sentiment analysis. So, for example, if you wanted to look at an email or the script from a call interaction with a customer service representative, you can look at it and kind of determine whether that was a positive interaction or a negative interaction in the organisation. And there are myriad tools for working with content, translations and transcriptions. I can take a video, for example, and extract the audio out of it and transform it into text. We actually use technologies like that to do subtitled videos. So a lot of great services out there. And, you know, typical modern content platform is going to provide you with a very intelligent and performant way to plug into those different services. Give you a kind of common framework for that integration. But we think more value comes from really working with your own content, your own data to train your own models. And so where we really see a difference in the market from an approach standpoint is giving people the tools that they need to go out and train, deploy, administer their own models. When you’re talking about Machine Learning and when you talk about really getting value out of this, I’ll give me a great example. We have a customer today. They work in the apparel industry. And over the course of every year, they generate thousands and thousands and thousands of photographs. So when you think about a photograph and when you ingest a photograph into a digital asset management system, the content services system. How do you find it? Well, that’s all the data associated with that photograph. That allows you to find it because it’s not really a self-described an object. So what we’ve been able to do with this apparel vendor is go ahead and train custom models for them that allow them to uniquely identify their own products, to identify the models or the talent that appear in these photographs, to identify what kind of shot it is, perhaps the context of the shot and really give them the tools. So as these new photographs come in from their different agencies and photographers in just a moment of environment and then automatically apply a bunch of data to them that allows that that those photos to now become findable, reusable and really valuable to the organisation from the standpoint of an asset management system for them. So a great example of how we can apply custom money. Couldn’t do this with a commodity model because commodity model can’t identify a particular Calvin Klein shirt. But if we train a customer custom model, then it has the ability to go out and do that. And this is where we see a lot more value for customers. We kind of call this business specific data, but modern content platforms should give you the tools and capabilities to begin to develop your own models as an organisation and apply those models and really get value from those models around your own products, your own services, your own customer experiences.
NICKY PENNYCOOK [00:14:31] Absolutely. That’s a really great example as well. And I think you’ve touched on it a bit there. And and it sounds like from my perspective, it creates a lot of ease within business. But what would you say are the advances and business benefits to organisers- organisations of using A.I. in Content Management and perhaps that they’re considering that move?
CHRIS MCLAUGHLIN [00:14:55] Yeah. So, yeah, we did touch on a little bit earlier. First and foremost, one of the real conversations we’re having with customers today, and it’s much bigger than just GDPR. But it’s a great example, right? If I go to an organisation today and I say, I want to know what information you have about me as a consumer, and we think about the fact that that information exists. And we just did a survey on this, right. For U.K. banks and insurers, for example, on average, nine different systems. Right. So if I am that bank, I have to look. And by the way, 8 out of 10 respondents that those systems are plugged together. Right. So that information exists in different silos. In order for me to be able to respond to a customer in what they call a subject action request. I have to go into nine different systems to kind of pull them together. So what we’re thinking about now is, one, how do we begin to plug these different systems together? You got to remember, each system has its own data model, its own way of describing information. So we need to think about normalising across those different silos and really getting a singular view into all of this customer information. Great example of where we can begin to use technologies like Machine Learning to enrich those different repositories, get kind of a master metadata model and make it easier for folks to access that information. So a great example for clients standpoint but I’m going to think about it also from a customer self-service standpoint or “hey, I want to deliver that information into a mobile application” or “hey, I’m a call centre operator and I want to see all of that information in one place be able to access exactly the same information that the customer has themselves in terms of maybe what’s been sent to them as statements or reminders or notices or other things like that”. So bringing all that information together has a tremendous advantage to organisations from a customer experience standpoint. What does that self-service or service that’s delivered by people organisation from an automation standpoint, we kind of talked about that forms process in use case earlier. There are fantastic opportunities for organisations to accomplish cost savings by bringing greater automation in to kind of traditional paper based processes. We were talking to one of our U.S. insurance customers and today, still today, 60% of the forms that they get into the organisation are paper based and they’re handwritten. So that’s very labour intensive. It’s very low value labour for an organisation. And it’s something that we can apply Machine Learning technologies to first to kind of say, “hey, have we completed the form correctly? Or is there something that we need to do?” But secondarily, then, we’ve been getting to extract the customer information from the form, perhaps kick off a new account onboarding process or a new policy process or whatever that is. Traditionally, we’ve kind of looked at scanning operations and things like that for this type of activity. But the reality is more and more, this information is coming through a variety of different channels. And what organisations are really looking to set up is kind of a Machine Learning service that can address that information regardless of where it comes from. The bigger thing for us and what and this is what I touched on earlier from an insight standpoint, a great example. U.S. auto insurer, one of the things they want to be able to do is ultimately begin to pull together a collection of, if you think about an accident nowadays and claim. In an automobile accident now you take a picture with your iPhone and it’s no longer and then fill out a police work or submit an accident report. I’ve got a photo. I might have a video of the accident. I might even have audio from different testimonials and things like that. So we’re work with all sorts of new content types as part of that process. And first and foremost, organisations are kind of trying to figure out how do we make it easier for customers to submit that information. But then secondarily, what this organisation was looking to do was begin to build a content lake with all of that accident photo information. You know, and if you think about different makes models of automobiles, you think about different types of damage. Excuse me. And then you think about the corresponding estimates and costs associated with that damage. Now we begin to have real intelligence about how those different accidents come together from an organisational standpoint. So now I can look at a photo, I can compare to previous photos, and I can think about then applying automatic estimates based on the information, the photos that I’ve received from the customer. So new insights into the business. But when you think about that longer term, that information really is a source of competitiveness, makes it more efficient for them to process claims. Gives it more intelligence about those accidents, about what they’re receiving from their customers. And ultimately, that information becomes kind of the foundation for business transformation.
NICKY PENNYCOOK [00:20:14] That’s really interesting. I think I can hear you absolutely fine, Chris. I think you’ve frozen a bit but we will carry on. We can hear you so thats the main thing.
CHRIS MCLAUGHLIN [00:20:25] I’m sorry about that and I see I froze in a very attractable pose, by the way.
NICKY PENNYCOOK [00:20:30] That’s okay, Internet connections we know that and you’re across the globe from our state. And say you’ve touched on a couple of different examples, which gives a little different insights, different industries but are there… do you think there are certain industries that are best suited in A.I. and Content Management. And we’ll say why you think not?
CHRIS MCLAUGHLIN [00:20:53] So, again, short answer, no. Longer answer, so one, opportunities in all sorts of different industries. Right. These challenges that I’m talking about and I know I’ve used, you know, certainly consumer products, retail is example. And I’ve also used financial services example. And those are industries we really focus on as an organisation. So I’m more familiar with the newscast.
NICKY PENNYCOOK [00:21:19] Sorry. I think we’ve lost your audio.
CHRIS MCLAUGHLIN [00:21:21] Oh, oh.
NICKY PENNYCOOK [00:21:25] I can hear you but I’m not sure “if we can had it on the output as well”.
CHRIS MCLAUGHLIN [00:21:31] All right. Can you hear me or…
NICKY PENNYCOOK [00:21:36] No, I think we’ve lost you. Is it- really sorry Chris but is it possible Chris, that you could leave and rejoin and return it to be? So are everyone. We are live and we are having a few issues. And Chris is recovering some really, really important points there. And hopefully we’ll be able to get him back to hear his insights on how certain industries are in using A.I. whether they should be better “sit” to A.I. And I think in the meantime, it’s a really important conversation to be having at the moment where things are changing. We’re adapting, different businesses having to make those adaptions and perhaps consider things like A.I. in business and Machine Learning. And hopefully I think we might have Chris back now. Possibly? Hello Chris? I can’t hear you. We’re having in a second. I’m really sorry about this. You think do we have Chris?Thank you. Oh, there we go. I think we might have you back.
CHRIS MCLAUGHLIN [00:23:24] Here we go.
NICKY PENNYCOOK [00:23:25] Yes. Sorry about that.
CHRIS MCLAUGHLIN [00:23:28] No, no, I’m sorry. I don’t know what the connectivity is or not, but I got stucked backstage there.
NICKY PENNYCOOK [00:23:33] All right. We’ve got you know there. So, let’s give you a little bit reminder… so we’re talking about whether there are set of industries are better suited to use an A.I. and contact management? And if you agree or disagree with that? And why that would be?
CHRIS MCLAUGHLIN [00:23:48] So, I don’t know that there’s necessarily one industry that’s better suited than another. But as an organisation, we really focus on financial services, particularly when we’re thinking about enterprise Content Management and content services. So we work a lot with banks, insurers, health care companies. And then for digital asset management, we work a lot with consumer products companies, retail companies. So those are my examples and I’m more familiar with use cases there. But the reality is, when you think about the challenges that we’ve talked about previously, they exist across industries. So if you’re looking at energy companies who are building power plants or delivering billing statements to their customers, they have the same challenges from an automation customer experience standpoint. Certainly pharmaceutical companies, life sciences companies, we’re doing clinical trials. Same same kind of challenges. So we really see a lot of opportunity for the technology across industries. I just happened to be more familiar on the financial services and kind of the retail CPG side.
NICKY PENNYCOOK [00:24:59] So just to sum up and look to future. What does the future of A.I. and Content Management look like? And what do you think enterprises will be doing in five years? Obviously, we’ve probably been for a bit of a cad food for this year where probably things are going at the moment. And into your perspective, how do you think things are going to move forward?
CHRIS MCLAUGHLIN [00:25:22] Yeah. So five years in technology is an incredibly long time, right? So all a little bit a little bit closer and a little less far out in terms of the time horizons. But we think from an Artificial Intelligence standpoint, we’re just scratching the surface here. You know, we talked about some fundamental use cases on extraction, on the animation and a little further out in terms of new insights into the business and content lakes. But ultimately, and when we think longer term, we kind of think about how we work with content, we also think about how we drive business processes with content. So some opportunities, you know, that we see further out really have to do with. And I think I touched on this earlier, kind of predicted delivery of information. We feel strongly that we can apply Artificial Intelligence technologies to begin to understand what information has more value to an organisation, what information has greater affinity to certain types of workers. So who who’ve used this information? So thinking beyond what’s in the document, but how the document is used. And as we begin to understand that, then I can say, “hey, this is the type of information that Nicky needs to do her job” and begin to push that information to you as you are performing tasks. So this is what we kind of mean when I talked earlier about contextual and being more contextual with information in the context of the work that you do, the role you perform in the organisation. So instead of kind of passively saying, am I to make it easier for you to find and retrieve information? Now I’m going to say, hey, I’m going to push information to you when, where and how you need that information. The other thing that we kind of look at is on the process side of the equation. If you look at it again or use financial services example, but large organisations who are processing claims and policies, things of that nature, there’s a lot of repetitive decision making, a lot of repetitive work in there. And one, we believe we can begin to automate some of that decision making. So instead of having maybe a customer service representative or a claims adjuster touch every single transaction, we can begin to intelligently identify exceptions that need to go to kind of a second level and a knowledge worker who has specific expertise, but for basic decisions that need certain parameters, we think we can completely automate that. Again, it brings costs and efficiency benefits to an organisation and really allows them to apply their knowledge workers more intelligently to their processes. The other thing that we kind of think about is learning business processes. So, you know, if, for example, at a certain step of the process, we always have an exception or we always push work to a certain work performer who has specific expertise. We believe that the process itself can begin to learn that behaviour can adapt to that and can actually adjust itself to make sure that we are as efficiently as possible routeing work to the right people to perform that work. So really kind of thinking processes. So, you know, hopefully I’m giving you a picture that we kind of think that the overall opportunity for Artificial Intelligence is unlimited. And it’s, again, not about displacing work. It’s really about making work more intelligent, about leveraging human resources more effectively in an organisation and really employing human resource where it’s most needed in a business process.
NICKY PENNYCOOK [00:29:07] Absolutely. Thank you, Chris. You’ve provided a really great insight, it sounds like there’s a lot more to come of A.I. and Machine Learning. As of now and as we move forward nto the future. And you’ve been a great guest. Thank you so much for joining me today. I’m very sorry about the technical issues.
CHRIS MCLAUGHLIN [00:29:24] Ah, I’m sorry about the techinical issues. Thank you for having me.
NICKY PENNYCOOK [00:29:28] That’s alright, you’ve been great. Thank you so much, Chris. Great. And so that was Chris and I’m from Nuxeo. It gave us some great insights into A.I. and in Content Management and also a bit of Machine Learning too. And so tomorrow, we’ll be back at 12:30, I believe. And that’s everything for today. If you enjoyed the show, if you were interested in taking part, visit us on Disruptive Live. Go check out our social media channels on Disruptive Live. Thank you, bye.