 This is George Gilbert. We're on the ground at demand base the B2B CRM company based on AI one of a very special company that's got some really unique technology We have the privilege to be with Seth Meyers today senior data scientist and resident wizard and who's going to take us on a journey Through some of the technology demand base is built on and some of the technology coming down the road So Seth welcome. Thank you very much for having me so We talked earlier with Aman Neymar senior VP technology and we talked about some of the functionality in in demand base and how It's very flexible and and reactive and adaptive in helping guide or React to a customer's journey Through the buying buying process Tell us about what that journey might look like how it's different and You know the touch points and the participants and and then how how your technology? Rationalizes that because we know old CRM Packages will really just you know Lists of contact points, right? So this is something very different, right? How's it work? Yeah, absolutely So I mean at the highest level each customer is going to be different Each customer is going to make decisions and look at different marketing collateral and respond to different marketing collateral in different ways You know as companies get bigger and their products offering become more sophisticated That's certainly in the case and also, you know sales cycles take a long time You know you you're engaged with an opportunity over many months And so you know there's a lot of touch points. There's a lot of planning that has to be done So that actually offers a huge opportunity to be solved with AI Especially in light of recent developments in this thing called reinforcement learning So reinforcement learning is basically machine learning that can think strategically They can actually plan ahead in a series of decisions and it's actually technology behind Alpha go Which is that the the Google technology that beat the best go players in the world? And what we basically do is we say okay if we understand If you understand your customer, we understand the the company work out We understand the things they've been researching elsewhere on third-party sites, then we can actually start to predict about content They will be likely to engage with but more importantly we can start to Predict content or more clearly to engage with next and after that and after that and after that And so what our technology does is it looks at all possible paths that your that your Potential customer could take all the different content He could ever suggest them all the different routes They will take and it looks at the ones that are likely to they're likely to follow But also ones are likely to turn them into an opportunity And so we basically in the same way Google Maps considers all possible routes to get you from your office to home We do the same and we choose the one that's most likely to convert the opportunity the same way Google chooses the quickest road home Okay, this is really it's a that's a great example because people can picture that but how do you How do you know what's the best path? Is it based on learning from previous journeys from customers? Yes, and then if you make a wrong guess you sort of penalize the engine and say pick the next best What you thought was the next best path Absolutely. Yeah, so so basically the way the nuts and bolts of how it works is we you know We start working with our clients and they have you know They have all this data of different customers and how they've engaged with different pieces of content and throughout their journey And so the machine learning model what it's really doing at any moment in time given any customer in any stage of the opportunity That they find themselves in it says what piece of content are they likely to engage with next? What and that's based on historical training data if you will and then once we make that decision on a step-by-step Basis then we kind of extrapolate and we basically say okay if they if they if we showed them this page or if they engage with this Material what would that do what situation we find them in the next step? And then what would we recommend from there and then from there and then from there? And so it's really kind of learning the right move to make at each time and then extrapolating that all the way to the opportunity being closed the picture that's in my mind is like The the deep blue. I think it was chess. You know where would map out all the potential moves very similar game very similar idea, so What about if if you're trying to engage with a customer across different channels and it's not just web content Yes, how is that done? Well, that's something that we're very excited about and that's something that we're that we're currently Really starting to you know devote resources to right now. We already have a product live that's focused on web content specifically But yeah, we're working on kind of a multi-channel type solution and we're all pretty excited about it Okay, so obviously you can't talk too much about it. Can you tell us sort of what channels that might touch? I? Might have to play my cars look close to my chest on this one, but I'll just say we're excited All right. Well, I guess that means I'll have to come back Please please so So tell us about that the the personalized Conversations is the conversation just another way of saying this is how we're personalizing the journey Or is there more to it than that? Yeah, it really is about personalizing the journey, right? Like you know a lot of a lot of our clients now have a lot of sophisticated marketing collateral And they have a lot of a lot of time and energy has gone into developing content The different people find engaging that kind of positions products towards pain points and all that stuff And so really there's so much low-hanging fruit by just Organizing and leveraging all this material and actually forming the conversation through a series of journeys through that material okay, so Aman was telling us earlier that that you know We have so many sort of algorithms that they're all open-source or they're all published And they're only as good as the data you can apply them to right so tell us Where do companies? startups You know not the Google's Microsoft's Amazon's where do they get their proprietary? Information is it that you have? algorithms that now are so advanced that you can refine raw information into proprietary information that others don't have or I mean really I think comes down to like our competitive advantage I think is largely in the source of our data and so Yes, you know you can build more and more sophisticated algorithms But again if you're starting the public data set you'll be able to drive some insights But there will always be a path to those data sets for say a competitor for example, we're currently tracking about 700 billion web interactions a year and then we're also able to attribute those web interactions to You know companies mean the employees at those companies involved in those web interactions And so that's able to give us an insight that no amount of public data processing Whatever really be able to achieve how do you Aman was was Started to talk to us about how like there were DNS reverse DNS registries or something reverse IP look-ups. Yeah. Yeah, so is How are those? If they're individuals within companies and then the companies themselves How do you identify them reliably? right, so reverse IP look-up is we but so we've been doing this for years now and so we've kind of developed a Multi-store solution, so you know reverse IP look-ups is a big one Also machine learning, you know, you can look at a traffic coming from an IP address and you can start to make some very You know informed decisions about what they what the IP address is actually doing who they are And so if you're looking at at the account level, which is what we're tracking at It's yes, there's a lot of information to be gleaned from sort of the way and This may be a weird sounding analogy But the the way a virus or some piece of malware has a signature in terms of its behavior You find signatures in terms of users Associated with an IP address. Yeah, and we certainly don't De-anonymize individual users, but if we're looking at things at the account level Then you know, you know bigger the data that the more single you can refer And so if you're looking at a company-wide usage of an IP address Then you can start to make some very educated guesses as to who that company is the things that they're researching What they're in market for that type of thing and and How do you know? How do you how do you find out? You know if if they're not coming to your site and they're not coming to one of your customer sites How do you find out? What you know what they're touching right? I mean like I can't really go into too much detail But a lot of it comes from working with publishers and you know a lot of the state is just raw and it's only because we can identify The companies behind these IP addresses and that we're actually able to Actually turn these webinar actions into insights about specific companies sort of like how like Advertisers or publishers would track Visitors across many many sites by having agreements. Yes, okay. Yeah along those lines. Yeah, okay, okay so Tell us a little more about Natural language processing I think Where most people have assumed or have become familiar with it isn't with the B2C Capabilities with the big internet giants where they're trying to understand all language. Yes, you have a more well scoped problem Yes, that changes your approach so so a lot of really exciting things are happening in natural is processing in general research and and Right now in general, it's being measured against this yardstick of can it understand language is better as good as a human can Obviously we're not there yet But that doesn't necessarily mean you can't derive a lot of meaningful insights from it and the way we're able to do that is Instead of trying to understand all of human language. Let's understand very specific language associated with the things that we're trying to learn So obviously we're a B2B marketing company. So it's very important us to understand What companies are, you know are investing in other companies what companies are buying from other companies what companies are suing other companies? And so if we if we said, okay We only want to be able to infer a competitive relationship between two businesses in a national document That becomes a much more solvable and manageable problem as opposed to let's understand all of human language And so we actually started off with these kind of open-source solutions with some of these proprietary solutions that you know We paid for and they didn't work because their scope was this broad and so we said okay We can do better by just focusing in on the types of insights. We're trying to learn and then work backwards from them So tell us how much of the algorithms that we would call built You know building blocks for what you're doing and others How much those are all published or open source and and then how much is your secret sauce? Because we talked about data being key part of the secret sauce. What about the algorithm? I mean, yeah, you can treat the algorithms as tools but you know A bag of tools a product does not make right so it really comes our secret sauce becomes how we use these tools How we deploy them and the data sets we put them again So as mentioned before, you know, we're not trying to understand all of human language Actually the exact opposite So we actually have a single machine learning algorithm that all it does is it learns to recognize when Amazon the company is being mentioned in a document So and so if you see the word Amazon is it talking about the river is it talking about the company So we have a classifier that all it does is it fires whenever Amazon's being mentioned a document and that's a much easier problem to solve than Understanding, you know than Siri basically Okay. Yeah, I still get rather irritated with Siri So let's let's talk about Broadly this topic that sort of everyone lays claim to as their great, you know higher calling which is democratizing You know machine learning and AI and opening it up to a much greater audience Help set some context just the way you did by saying hey if we narrow the scope of a problem It's easier to solve. What are some of the different approaches people are taking? to that problem and and What are their sweet spots, right? So, I mean the kind of the talk of the data science community talk in machine learning right now is some of the work that's coming in of DeepMind, which is a subsidiary of Google they just built AlphaGo which you know solved a strategy game that we thought we were decades away from actually solving and their approach of restricting the problem to a game to with well-defined rules with a limited scope I think that's how they're able to propel the field forward so Significantly they start off by playing Atari games, then they move to you know long-term strategy games and now they're doing video games Like video strategy games And I think the idea of again narrowing the scope to well-defined rules and well-defined limited settings It's how we're actually they're actually able to advance the field. Let me ask just about the playing the video games And I can't remember a star Starcraft Starcraft would would you call that like? where The video game is a model and You're training a model against that other model. So it's almost like they're interacting with each other, right? So so it really comes down you can think of it as Pulling levers right so you have a very complex machine and there's certain levers you can pull and the machine will respond in different ways You know if you're trying to for example Build a robot that can walk amongst a factory and pick out boxes like there's how you move each joint how you you know Where you look around all the different things you can see and sense Those are all levers to pull and that gets very complicated very quickly But if you narrow it down to okay, there's certain places on the screen I can click There's certain things I can do there's certain inputs I can provide to the video game You basically limit the number of levers and then Optimizing and learning how to work those levers is a much more scoped and reasonable problems opposed to learn everything all at once Okay That's interesting now Let me switch gears a little bit and you know We've done a lot of work at at wiki bun about IOT an edge, you know increasingly edge-based intelligence because you can't go back to the cloud You know for your analytics for everything, but one of the things that's becoming apparent is It's not just the training that might go on in the cloud. Yes, but there might be simulations You know and then the sort of low latency response is based on a model. That's at the edge Yeah, help elaborate what that really where that applies and how that works, right? Well in general When you know when you're working with machine learning in almost every situation Training the model is that's really the data intensive process That requires a lot of extensive Computation and that's something that makes sense to have localized a single location, which you can leverage resources and you can optimize it Then you can say all right now They have this model that understands the problem that's trained It becomes a much simpler endeavor to to basically put that as close to the you know That's the device as possible and so that really is is how they're able to say okay Let's take this really complicated. You know billion parameter neural network that took days and weeks to train And let's and let's actually derive insights at the level right at the device level Recent technology though like I mentioned deep learning that in itself just the actual deploying the technology Creates new challenges as well to the point that actually Google invented a new type of chip to just run Yeah, the TPU the tensor processing unit just to handle what is now a machine learning algorithm so sophisticated that even deploying it After it's been trained is still challenge. Is there a difference in the in the hardware that you need for for training versus inferencing so they They initially deployed the TPU for just for the sake of inference in general The way the way it actually works is that when you're building a neural network There's a type of mathematical operation to do a whole bunch and it's it's basically working with matrices Yeah, it's like that that's still absolutely the case with training as well as inference Actually, you know querying the model But they yeah, so if you can solve that one mathematical operation then you can deploy it everywhere. Okay. Yeah so if if One of our our CTO was talking about how in his view What's gonna happen in the cloud is? richer and richer Simulations and that they as you say the querying the model getting an answer, you know in real-time or near real-time is that on the edge What exactly is the role of the simulation? Is that just a? Model that understands time and not just time but many multiple parameters that it's playing with right So so simulations are particularly important and taking us back to reinforcement learning where you basically have many Decisions to make before you actually see some sort of you know desirable or undesirable outcome And so for example the way alpha go trained itself is basically by running simulations of the game being played against itself And really what that simulations are doing is allowing the artificial intelligence to explore the entire Possibilities of all games sort of like war games. Yes, remember that movie. Yes With that yeah, Matthew Broderick and it actually showed all those war game scenarios on the screens Yes, and then figured out you couldn't really win, right? Yes. Yeah It's a similar idea where they and because for example and go there's more board configurations There are atoms in the observable universe And so you like you know the way the way deeply one chess is that basically more or less explore the vast majority of Chess moves You can't that's really not the same option You can't really play that same strategy for the alpha go and so this this constant simulations how they explored the meaningful Game configurations that it needed to win so in other words they were rather they were scoped down So the the problem space was smaller, right? And in fact basically the one of the reasons like alpha go is really kind of two different artificial intelligence working together One that decided which Solutions to explore like which possibilities it should pursue more and which one's not to to ignore and then the second piece was okay Given a certain board configuration What's likely outcomes and so those two working in concert one that narrows and focuses and one that comes up the answer Given that focus is how I was actually able to work so well. Okay, yeah That's on that note. That was a very very enlightening 20 minutes And I'm glad to hear that we'll have to come back and get an update from you soon All right, absolutely. This is George Gilbert. I'm with Seth Meyers Senior data scientist at demand base a company I expect we'll be hearing a lot more about and we're on the ground and We'll be back shortly