hello everyone and welcome to ask the expert session where we'll talk about the science of engagement personalizing customer experience with behavioral science and data driven inside my name is asrar and i'll be the host for the session i myself from the microsoft startups team leading the business development and sales engagements with our startups joining me today is our esteemed guest corey scobee who is the chief technology officer at cement corey why don't you start with telling us a little bit about yourself thanks azran and thanks for having me here it's great to be here so as you mentioned i'm the chief technology officer at cement i'm actually relatively new to the organization i just joined the company in uh in january but man's been doing some pretty amazing things for the past several years we've been building a platform that i'm eager to share some details about around behavioral science and analytics and how that engage how that correlates to customer engagement great thank you and as as we deep dive into the session i also want to let everyone know about the microsoft code of conduct essentially we would really appreci we want to provide a safe and sound and healthy environment for all of the participants so we really appreciate uh if you can go back to the previous slide where we can actually show the microsoft code of conduct and let the let the audience go over it for for a quick second um so really at the end of the day with microsoft does not really uh tolerate any harassment or discrimination we want audience to like i said we want everyone to have a safe and respectful and engagement throughout the session and if you have a question please type it in the iron window that you can see on the right side of the screen and if there's a question that the audience really likes please hit the like and our moderators will surface that question over to us uh so with that i guess let's go ahead and and get started uh so so corey why don't you tell us a little bit about cement and provide a brief overview about what you do and how you're helping the customers yeah absolutely would love to so cement is a technology company and really we're focused on helping other companies maintain better relationships with their customers often customers um fall either into a category of not being in good standing with relationship to their financial profile with their with with their vendors right we're talking about like telcos and utilities and that kind of thing or maybe they may otherwise be financially at risk for other reasons and so really what we're doing is we're combining behavioral science and computer science to analyze and automate customer engagements and to help nurture those customers back to health so at the end of the day what samanda's a company does is that we operate a sas platform for engagement and analytics that supports this kind of automated engagement with customers um on behalf of you know mostly uh larger but also mid-sized companies that have some kind of a month-over-month or week-over-week financial relationship with their clients that's really great to hear by the way i also want to congratulate you on raising series b extension recently uh can you talk to us a little bit briefly just so the audiences are here a little bit about your journey so far especially pre-pandemic as we went into the pandemic and the explosive growth that you have seen and you are on the trajectory to continue seeing in the space yeah you know it's been a really exciting a couple of years for cement the company was actually founded about four years ago by a pair of co-founders who really believe in the mission of of using behavioral science and and human empathy to help engage customers and and build better relationships between customers and companies and the past couple of years have been like a very uh steep growth curve for us so as you mentioned we we signed a series b extension um at the beginning of this year which was an extension to the original series b that we did just nine months ago so in the last 12 months the company has actually raised over 100 million dollars in venture capital in order to fuel this and we're seeing just tremendous amount of market demand on the other side so we're engaged with some of the largest telcos in in north america with some of the largest financial institutions and our customer base is growing very rapidly and cement as a company has more than quadrupled in size in the past 12 months as well so that's that's incredible and we are aware of of all the the specific industry like telcos being just one of the the many examples where you're seeing significant growth i want to touch a little bit about the platform architecture now i we know you're one of the strategic partners of the microsoft for startups program and the program is really focused on the technology as well as the sales and enablement now since the audience is very technical in nature i want to deep dive a little bit about the on the technology side can you talk a little bit about the arc the your your platform architecture and how cement uniquely tackles the customer problem in this unique space thank you yeah for sure um so we are in fact a heavy user of azure uh cloud services public cloud services across across a wide array of different services available on the azure platform we chose azure and the cloud because as a hyperscale company we're looking for the reliability and the growth and the scale that comes with being a microsoft customer fundamentally at its foundation what cement does is we take information from our customers from the companies that that have a relationship with clients we take records about what state um people are in and and if there are people that are are at risk or delinquent from their their current bills and we take large amounts of that information into the platform to help build a behavioral profile for individual users and then to automate the engagement of outreach to them and in many cases outreach is uh are through channels such as sms or email automation um etc and giving customers options so if you look at the foundation of what we're using um we have a huge data ingest service which um is in in largely built on adf today which is azure data factory um we are a heavy user of sql server um across the various different rdbms facets of our platform we're heavily into azure functions as well so a lot of our processing or calculation is done at scale across a service bus with azure functions we use azure web apps to deploy uh landing pages which are the the calls to action that our customers receive as part of the output um and we're also working with some microsoft partners as well so we you know we're working with snowflake um on azure as an example for uh data for data lake services um in analytics etc so that that's really amazing thank you for sharing that you gave us a lot more on the azure side i really want to pivot for a second towards this towards the startup side and and you being a cto of a growing company as a cto what would your advice be to other startups we're in the process of making this technology decision which is one of one of the most important decisions that a startup would be making or a company in general yeah you know so it's interesting as you um go through the different phases of building a company you typically go through a series of phases that that happen fairly quickly so there's the i have an idea about some technology that would help the world and i need to prove that that idea is valid phase in which case you write a little software to do it and you see how it how it kind of works out um and if you get traction with that you very quickly move into the phase that that i think cemented a lot of other companies are which is the growth phase of that and at the growth phase you have a there's some really seminal decisions to make about the platform technology you choose to use and how scalable and reliable that can be over time and my advice to ceo ctos and to ceos for that matter if i'm giving free advice would be build for the scale that you envision the business is going to be in the future and don't take shortcuts in the short term and so you know our use for example of azure functions is because we envision growing from where we're at today which is millions of transactions to pay potentially hundreds of millions or even billions of transactions a month on the platform and so we're always building for sort of the next phase of growth in our in in how we are constructing the solution on top of um on top of azure and the reason that that's so important is because if you do get the right formula and you do hit the right mark and you are a hyper growth company there's never any time to go back and fix it later on you always are moving forward forward forward and so my advice is you know is be pro be pragmatic but think big in terms of how you're thinking about the scale of your solution when you're picking cloud providers and when you're selecting technologies associated with that absolutely i just want to pivot real quick since you're build on azure 100 want to understand what are some of the benefits uh that microsoft has been able to provide from an azure standpoint from an engineering and technology standpoint um that you would like to like the audience to sort of think about and take away as they're engaging on that or thinking about or making that technology decision yeah so i you know ultimately um we you know part of the reason that we chose azure as a as a platform is because it aligned well with the skill set that we had at cement and and it was well-known technologies within the space as we as we think about um each evolution of our platform we're sort of thinking about all of the various different facets that are going to trip us up as a company who who deals in who brokers in other companies um customer information securities at the top of our mind right as a company who is a company who is required to take in large amounts of data process it on a on a very tight schedule on a daily basis to be able to pull um and manage that that kind of a data turnover scalability is on our mind and so you know when you add all of those things up i think uh our relationship with microsoft has been definitely fruitful one microsoft has been great in bringing resources to bear to help us understand best practices and patterns and implementation patterns that would make sense for us and so my my team engages with microsoft technology team on a regular basis to gut check our assumptions about how we are implementing the technology um and also there's you know inevitably it's software development so it doesn't always go perfectly and so uh whenever we have an issue and we are unable to find a root cause ourselves we call in our microsoft partners and you know together we can always find our way to the to the solution so abs absolutely thank you for that corey talking about software development i think one of the most important things is technology is the field for infrastructure it basically provides that underlying cover but that's something that's really important and i think this is where cement has done a phenomenal job is especially being in the customer engagement and customer experience space is empathy can you talk a little bit about the empathetic approach you're taking towards delivering these high quality customer experience starting from software development all the way to that end customer yeah so one of the things that we found early on in the development of cement was that um customers felt uh were engaged when they felt uh that like they had options and so part of the the whole modality of the behavioral science that we bring to bear is to try and make customers feel empowered to be able to make decisions on their own about their own outcomes and so you know what we've been doing from a software development perspective is building more and more modularity into the system that allows um a client a telco or a utility or a bank to be able to select or and create more and more paths that more closely match to the individual profile of their customers and then through analytics and and deep deep learning um and even some you know ai and ml types of things like we use uh nlp for example we use natural language processing to gauge what the what the um customer satisfaction um score or customer satisfaction feeling is around the engagement of of a customer when we go through the various different reach outs and engagements and then from there we can use that to tune specifically to a customer so really it's all you know the convergence of the human of the computer science and the empathy part of it is that we can use anchoring off of empathy we can use computer science to create more customized experiences for customers and that leads to empowerment and that leads to a feeling of loyalty between the customers and their and their clients and they're sorry they're they're vendor companies they're partners that's that's really helpful and and you know at the end of the day it's all data driven uh i want to take a quick pivot here because of the question that we came from from the audience so the question is as a product manager uh you know i need to we need to make data-driven decisions based on customer behavior which makes sense now corey in your experience have you ever found yourself in a role where there was little to very minimal data due to the nature of the business let's say dealing with government for example if so how would you then go about creating a data lake uh for such scenario yeah so i mean i think there's definitely places where we're even in our scenario we have blind spots about our customers and and some of it has to do with privacy and the and the reality of sort of privacy in the digital age right which is that as an individual customer um i may know things about you for a period of time like your name and your address and your telephone number but you have to be responsible in how you you use that so when it comes down to data and analytics on the other end of the process what we're really doing is trying to to find patterns and to build an analytics platform and a data lake that allows us to identify not specifics about an individual but the patterns that come from a certain set of demographic or or geographic markers for example and that can potentially help us segment and qualify and and treat customers in different ways so i guess my advice to the to the asker is um you use what data you have um and then you and if you can accumulate even if it's a small set of data points if you can accumulate enough time series of that data it's possible to use that to to identify pattern recognition and or leverage things like machine learning that can identify pattern recognition that we as humans maybe wouldn't recognize in a in a normal way that's very helpful thank you for that i want to also come back to all the exciting stuff that cement is working on before we talk about that i want to understand from your perspective being a cto you've seen a lot of success in some industries like telecom being one of them can briefly talk about your industry approach and uh and and really from an industry standpoint give the audience a sneak peek as to the industries where cement has seen significant growth and the industries that you're you're thinking of expanding in yeah for sure so um the original set of customers for cement was in fact uh from the telco market here in north america but what we found is that the idea of using behavioral analytics and customer and automating customer engagement is actually um applicable across many many different industries so the kinds of of industries that we typically see are industries where there's a uh a customer company relationship where they have where they're extending credit in some way shape or form and when we say extending credit it sounds a little weird because that sounds like a credit card or you know or a loan from a bank but when you're for example when your cable internet provider allows you to use their service for a month before you have to pay for it that is in fact extending credit so sort of the hallmark is companies that have some kind of a recurring financial relationship with their customers and are managing that relationship over time and then ultimately um you can use we can use the kinds of of technologies that we've created in the behavioral science platform and then the automated digital outreach platform in order to to manage those relationships so we see ourselves expanding um in many different vectors across industries so financial services utilities uh the banking industry in particular more and more of this subscription economy is driving towards a credit oriented approach and a recurring revenue kind of approach and so as more of our services as a daily consumer come into that that play cement has an opportunity to expand across that and then we also see ourselves expanding vertically as well so we have been focused primarily on customers on helping companies manage customers that go into what we call delinquency so or financial risk so a period where they're behind but they haven't actually cut off service and they're still managing a relationship with their customers we also see an opportunity to um to help in that post-delinquency scenario where a customer just cannot pay their bill and and really they need to move into some kind of a collections modality um and so that has historically been a pretty inhumane kind of industry and we think we can bring some empathy and some digital automation into that and then there's also the managing the customer relationship pre-delinquency as well which is you know who doesn't want to have a great relationship with their customers regardless of what what their financial status is um and so you know we think that the basics of the behavioral analytics and engagement platform that we're building also gives us the opportunity to expand earlier into the customer life cycle as well absolutely that's really helpful one thing that came to my mind as as you were talking about all these different things especially the touch point is now the face of a brand or these specific touch points where you know code happen and sometimes life takes turn for for the worse but what's really important working you know through cement brands are able to provide that differentiated experience through those touch points that are data driven number one but more importantly that are customized to that specific user and really very well curtailed to their specific scenario sort of providing that very unique and customized experiences that brands wish for so that's really great to hear another question we have from the audience sort of on that industry pivot is uh is around managing risk so by investing in technology it's at scale of quoting code where i want my business to be rather than where it is isn't there isn't this a particularly big risk as at present given the current pandemic market and the sort of unknown situations that are affecting many different industries so i think what my interpretation is how do you sort of balance that risk and sort of that unknown as you had as you know cement is thinking about a specific industry yeah i think that's really where the the power of you know the modern cloud-driven environment really comes to play which is that i can have my software development team write a web scale horizontally scalable implementation on top of on top of azure services and then i can scale the azure services as needed so i don't have to put a whole bunch of money up front into the volume that i think i'm going to need in 2023 by building a data center and a bun and pre um pre-provisioning a bunch of database instances and all that kind of stuff i can bring it online immediately with an agile devops kind of approach and so i i think the risk is is more mitigated than it used to be i think the idea that you can build a billion dollar company on top of a relatively low capital footprint um but you have the ability to scale beyond that at any point in time it is is one of the unique things about bill about building a startup in this age versus what it looked like you know 10 or 15 years ago which was a completely different thing i mean gosh if i if we went back in time 10 10 or 15 years ago i think there's a tremendous amount of risk to building the kind of scale that you envisioned for the future as a company built like i'd be building data centers and trying to understand how i'm going to operationalize on a global footprint and all that kind of stuff and boy it's just such a different world these days absolutely we'll take one last question from the audience before we we wrap up here in your machine learning approach do your data science teams work offline and develop models or tune features etc on local machines with quick iteration and then essentially another team figures out how to deploy these in production at scale in azure or does your data science team build these models right from the start in a cloud environment itself yeah so interestingly the answer is both right so we use data science in what i would describe as preceded model sort of algorithmic ways so nlp is a great example of that which is that we nlp is a real-time service so every time an email comes back from a customer we analyze it to figure out what the sentiment is inside of that email so the service that we operationalize there is one that is is actually a real-time service that's bound to our data processing system however the model for that is actually built and populated and repopulated offline over time and and you know similarly we have other scenarios where we are building um largely online in in and taking in real-time data so like our customer segmentation sort of um modality we think about that as more of a real time because we're funneling data off our mainline data processing through the models on a regular basis so we're building that kind of stuff in the cloud directly absolutely that that's really helpful i think i made this comment early on but i wanted to also come back to this about what's next for cement so as a cto i mean obviously you have insight into all the exciting innovation that we should expect from cement can you give the audience a sneak peek sort of what's coming next here yeah you know i mean i think we are continuing to expand our footprint of outreach so as we grow globally as a company different channels become more relevant in different geographies i'll give you an example here in north america it's pretty standard fair to use sms to to um to communicate with end customer users in different parts of the world it's primarily wechat or it's primarily you know different channels across different sort of of of global things so we continue to expand um our platform from a channels of outreach perspective um we're doing a lot of work with um again with data science and ai and ml to try and figure out how we can fine-tune the very specific experience that customers are are having in engagement pipelines um and yeah and just a lot of scale and growth i think you know we've got some big sort of plans under the covers architecturally that i won't be mean much to the world from a a public-facing perspective but you know is part of our building the next generation sas platform that's going to be the you know the definitive platform of customer engagement in this in this space absolutely i i also want to let the audience know that you know you can learn more about cement by visiting www.cemen.com and cement also has a has a solution on azure marketplace uh we also have some contact information uh for cement if you're interested in contacting the team and learning more or or running a pilot which we have seen a lot of enterprise customers as well as uh small companies um work with and or and or deploy and then essentially please follow uh please follow cement on linkedin twitter instagram and facebook now corey coming back to you back to you here in closing can you please tell us what can a business expect to achieve as a result of partnering with cement something that the audience should take away yeah i think really it gets back to what we were talking about with empathy and the customer relationship before which is that when you partner with cement and you use our automation platform what you get at the other end is you get a first you get really two effects one is that we help companies um manage and recover more of their customer uh revenue than they would otherwise right so there's a very hard dollar value to that which is if a certain percentage of your customers are going to go delinquent over time the faster you can heal them and the more that you can heal that relationship the you know the more financially beneficial it is but i think the real long-term value is the customers that go through and experience an event like that where they're in some series of delinquency whether it's through extenuating circumstances or just by accident but experience a really high-end customer engagement relationship become incredibly loyal customers downstream and we have a tremendous amount of data to show that so i think the real value is the just building the loyalty with your customer base it's we know how hard it is to obtain customers in in any market segment in this world and so keeping the ones that you have and making them happy and loyal um i think is really really important absolutely and then that directly ties to providing that brand experience that initial promise that you approach you know that you made to all the customers uh that you have so thank you so very much corey for being here and answering all the questions this has been a phenomenal session and the engagement from the audience has been really awesome again i want to thank you for taking the time out of your busy schedule and being at microsoft ignite with us thank you everyone and enjoy the rest of ignite thanks again thank you
2021-03-09