Find the PPT here: Reltio Data Product Customer 360
In this webinar, we will explore how Reltio's Customer 360 Data Product leverages AI to unify customer data, driving better customer experiences. Discover the benefits of a comprehensive and real-time 360-degree view of your customer data, powered by advanced entity resolution and AI-driven insights.
Key Highlights:
- AI-Driven Data Unification: Learn how Reltio uses AI to enhance customer data unification, ensuring a single, authoritative view of your customers.
- Real-Time Data Insights: Understand the importance of real-time data for personalized customer engagements and operational efficiency.
- Flexible Entity Resolution: Discover the power of Reltio's Flexible Entity Resolution Network (FERN) for accurate and efficient data matching.
- Generative AI for Data Exploration: Experience the capabilities of Reltio Intelligent Assistant (RIA) for conversational data interaction and advanced segmentation.
- Business Impact: See how a unified customer data product can drive growth, improve efficiency, and enhance compliance.
Whether you're looking to improve your customer service, personalize marketing efforts, or drive sales effectiveness, this webinar will provide you with valuable insights and practical examples.
Webinar Details:
- Host: Chris Detzel
- Guest Speaker: Venki Subramanian
- Topics Covered: AI in Customer Data, Real-Time Data Unification, Entity Resolution, Generative AI, Business Impact
Don't miss out on this opportunity to enhance your understanding of AI-driven customer data unification. Watch the recorded session to gain actionable insights and learn how Reltio's Customer 360 Data Product can transform your customer experience strategy.
Transcript below
Chris: [00:00:00] All right, why don't we go ahead and get started? I like to get started on time so that we can get through The entire presentation. So welcome to another Reltio community show. Most of you know me. I'm Chris Detzel. We have special guest Venki Subramanian. He's our senior vp of product Management at Reltio and he's done this before Venki.
Chris: How are you?
Venki: I'm doing very well, Chris. How are you?
Chris: Doing well. Really excited about this particular session. Reltio Customer 360 Data Product. Me too. Yeah, it's Powering AI Driven Data Unification for Enhanced CX. Now, before we do get started, everyone, keep yourself on mute. Questions could be asked in chat.
Chris: Or take yourself off mute and ask we'll probably get through the presentation and demo first, then we'll start asking questions. Don't want to interrupt that. It is being recorded for everyone to to watch later. My goal is to have that at the end of the day today or tomorrow. So upcoming events [00:01:00] today, we do have our customer 360 data product.
Chris: And then on the 22nd, we've got even new stuff coming out on our product release, customer and partner webinar. So make sure you tune in for that next week. I know I'm excited. And then we have a couple of shows. Around Reltio integration hub, one specifically address clean cleansing and techniques with Reltio integration hub.
Chris: And then one on advanced data enrichment with Reltio integration hub. I know that's a super popular. Area is our RAH. So we want to try to make sure we're getting more information out there to you and how to use it. So Venky, I'm going to give it over to you and let you share.
Venki: Thank you. Thanks. Thanks, Chris.
Venki: And like I said, I'm really excited to be here talking about this this topic of customer 360 and how AI augments customer 360 for providing better customer experiences. I'm going to share my screen in just a few seconds.[00:02:00]
Venki: And I think of this as a continuation of some of the recent conversations we've been having with you through our community shows, right? So we had in March. Community show around conversational experiences with relative intelligent assistant. Last month or maybe a couple of weeks ago, we had a conversation with you actually two different community shows on Pern on the flexible entity resolution network.
Venki: Customer 360 is where we are bringing all of these innovations together to enhance customer data, customer experiences ultimately for your end customers. So that's what we will be talking about today. Now, as part of this conversation, I will be talking about our existing functionality as well as some new capabilities that are being planned that are yet to be delivered.
Venki: So please make sure you understand, from that context. That's why I just wanted to make sure that we have our safe harbor statement in here. Now, before we dive into the into customer 360 and what specifically are we doing now and what we're doing differently [00:03:00] about it, I think it'll be good to take a moment to reflect on why we're talking about customer 360.
Venki: And first of all, I want to acknowledge that this conversation is not new at all, right? For many of you here who have been in the data management. In the customer data, customer experience space. Customer 360 is a term that emerged probably over a decade ago, and I can go back to, internet and do searches and find references to this that dates back about 12, 13, 14 years.
Venki: And the idea has always been about how can we bring together a unified view of our end customers. So that we can improve their experiences, right? So the reasons why custom, organizations invest in a customer 360 are to improve omni channel engagement for customers to personalize the customer journey, map the journey so that we can understand their journey through every step of the process so that we can influence it positively.
Venki: To provide seamless experiences, to improve internal operational [00:04:00] efficiency across business functions. And one of the examples that we will talk touch upon today is customer service. How many of you would have had experiences where you're calling into a contact center, let's say to ask about a claim on an insurance policy or to make a change to an insurance policy or simply to renew and just having to be handed over from one person to another and having to explain your situation or your needs all over again, right?
Venki: And those are examples where we have as customers seen the challenges that can that can exist in an organization if you don't have a unified consistent view of your customers. So operational efficiency is an important aspect, brand loyalty, retention, and growth, ultimately happier customers buy more, right?
Venki: So it is really simple. Why organizations care about it? Because there is a direct impact on your top line on revenue and growth. Okay. Which is highly correlated to happy, loyal customers with a high net promoter score. These are all different terms that you would have heard, and they're all correlated to, [00:05:00] again that customer experience end to end, and to impact that customer experience positively, you need to know your customers first.
Venki: And there is another aspect about data privacy, preference, consent management. Overall, if you think of compliance, and reducing risk with customer related engagements even in, let's say in financial services or in other related industries, think of processes like fraud detection or risk intelligence that also relies on having a consolidated, complete view of your customer.
Venki: So customer 360 is so foundational that for over decades, IT leaders and data leaders have been trying to invest and create a true customer 360 for for their organizations. But then why are we talking about it today? We are talking about it today simply because this has been a highly challenging journey for those customers.
Venki: And we still don't have a clear solution that has emerged out of the many different possibilities that exist. [00:06:00] And customers data landscape still look highly fragmented and siloed as you can see from this picture, right? Even for companies that have tried to create a customer 360, some companies have invested in a data lake house or a data warehouse and try to create a customer 360 there.
Venki: Some companies have invested in a. Customer data platform or other related technologies, maybe even legacy MDM technologies have been tried to use to create a customer 360 and they have, they all have their own specific challenges. The cloud data warehouse based approaches have high cost of implementations.
Venki: Because a lot of it is, it needs to be pulled together by combining different technologies. The customer data platforms by nature lack a multi domain view of the customer. So they can maybe understand the customer attributes, but if you want to relate that to products or subscriptions or other kinds of related information that are cost that you need to know in the context of a customer.
Venki: Those platforms are not suitable for that. And when you think of [00:07:00] customer 360 that resides inside applications, like a CRM application might have a customer 360 in there, right? So there are other applications that might solve it, but ultimately what is those add to these different silos. And even though there might be a cluster of information, Where within marketing, somebody might claim that I have a customer 360 that is again, unfortunately a siloed view of the information that only resides within a certain line of business.
Venki: So mature organizations obviously have tried to solve it. And one such approach that we have seen is by creating these use case specific data sets, right? So if you, if there is a specific initiative in an organization, let's say we want to improve cross sell or we want to improve our churn and protect or reduce churn, those kinds of business initiatives will then drive creation of a specific data set, which supports that specific initiative.
Venki: And these use case specific data sets. only adds to the problem because these [00:08:00] use case specific data sets further deepen the silo problem by creating copies of data. A lot of the same data is being replicated, copied over, duplicated even. And what it results is the core data for each domain gets reworked, which leads to poor quality data pipelines become fragmented.
Venki: Each use case potentially uses a different tech, adding to expense and complexity, more resources, more time, and they become very brittle in nature, right? When you want to change something, when the needs of the business change, then it becomes really hard to make those changes in a consistent manner across that impacts multiple of these different outcomes.
Venki: So this old approach has given way to a more modern thinking around data products. What is the market view of data products? And here is a slide from McKinsey that talks about a new way of solving these problems. Instead of solving them for specific use cases, how can we look at the core data being unified as a [00:09:00] product and different analytical and operational use cases being enabled on top of that product, as you can see up here.
Venki: And the advantage of doing this is as you add incremental use cases or scenarios on top of the data product, The incremental cost, the cost of adding or supporting yet another use case, another requirement incrementally becomes less because you're not duplicating everything. You're not trying to create everything from scratch, but you're essentially adding onto something that already exists.
Venki: This is typically what we do in the digital world as product managers, right? I've been in the product management space where every single digital product that I build, it has. incremental functionality that gets delivered. And that is what essentially comes out as a roadmap and you execute on the roadmap where you're not trying to build new products for every new functionality, but you're essentially adding those features to the product.
Venki: How can we apply the same principles to data? So that is why there is, I would argue there is a [00:10:00] resurgence of this 360 and truly thinking of it as a product. As a data product that organizations will need to create for them to power their omni channel experiences or any other outcomes that you're targeting with customer data.
Venki: So this is again interesting because this is not, as you can see, this is not just a relative point of view. This is something that the whole market is orienting itself around. And even in this example that McKinsey talks about, though, the 1 of the main examples of 1 of the leading example that is on the slide that talks about a data product is customer 360.
Venki: So that's a lot of context, right? There are a lot of reasons why this is important, why we are talking about it. What is different? What is changing in the technology as well as the data management landscape overall, that is making us take a different look at solving the same problem, solving the problem of customer 360.
Venki: And that is why we have combined the notion of that customer 360 and data product. [00:11:00] And really coming out with a new approach of solving a customer 360. The legacy approaches, like I said, I talked about earlier about using data warehouses or using a specific platforms or even applications.
Venki: They have had their challenges in the past. You can see how data moves from its source to the different consumers. And the consumers of a customer 360 are typically applications, digital automation, self service kind of experiences. And for that, the foundational customer data needs to be unified, deduplicated, enriched, so that there is a single authoritative view of who your customer is.
Venki: And then you have to augment that with all the interactions or transactions or events. that give meaning to to that customer's experience. And all of this information used to get combined in a data warehouse. And then you have to bring your AI data science kind of capabilities to that model to augment it with certain insights, right?
Venki: Even if it is something as simple as if you want to categorize customers by, into different [00:12:00] segments based on a certain propensity to buy you will need to apply a data science model in a data warehouse, and then you will extract and. You can activate that essentially activation in this case means can I extract my my early adopter customers out of a segment?
Venki: And then can I target them with a campaign? So these kind of very batch analytics oriented kind of processes are being done to a large extent by organizations today by leveraging a data warehouse. But when you need the same information at the point of engagement, going back to that example of a contact center that I talked about when the call center agent needs to pick up a phone.
Venki: And he or she needs to know that the customer that they are talking to is an early adopter and they would probably buy immediately provided some incentive like a coupon or a discount or something that information needs to be available at the point of engagement, right? Not in some analytical view, not in some data warehouse, but in the application at the point of engagement.
Venki: So how do we solve that? And that is one of the challenges that these these approaches [00:13:00] have. So what if there is a newer approach, a new differentiated approach using relative customer 360 data product, which can combine your customer data, the entity information, the interactions that are associated with that customer, whether it's a a web visit, but I went to a website, downloaded a white paper, or I actually executed a transaction.
Venki: I placed an order. Or whether I contacted the contact center with an issue that I might have, what if all of that information is being brought together into a single centralized unified view of the customer and that information that then available in real time for points of engagement, right? Whether it's a self service application, whether it's a contact center agent interacting with that customer, whether it's sales executive Interacting with the customer to help them complete the purchase or any one of those scenarios.
Venki: What relative customer 360 data [00:14:00] product enables is that single unified view of the data available in real time. Not just that, what if we are able to bring AI models to that data instead of taking the data to different systems and applying models and bringing results back? What if we are able to bring AI models to that data and augment that with derived insights, predictive insights, and that information is also now available as part of the same profile so that it becomes actionable.
Venki: You can take action on all of that information. And what of this information is available in your cloud data warehouses to support any other analytics processes, analytical workloads that you need to run. And all of this information in a unified manner is being distributed to your. Digital automations, your application systems, your B.
Venki: I. Analytics and including the newer Jenny, I investments that you're making all of that can rely on the single trusted view of your customer data. The 3 60 view of your customer data. So this is the new or different approach that we're talking about when we talk [00:15:00] about customer 3 60 data product as a new offering from relative.
Venki: So let's talk about what benefits this could create. By creating that single view of the customer, this becomes truly a reusable data asset with interactions, derived attributes, and more. Segmentation and activation capabilities built into it so that now you can power your consuming systems, your operational systems, your AI machine learning systems, all your analytical applications with that data and the benefits of you of using of building this on the proven platform of relative is you have high data quality at faster speed.
Venki: Essentially, we are a real time oriented platform. So you're getting all of the benefits of that. These data assets now are now reusable across different use cases. You can power multiple different call it activation scenarios with the same foundational data that is a simplified architecture with high degree [00:16:00] of business flexibility, which means if you want to make changes to the kind of information that you're capturing or the kind of information you're making available to your consuming systems.
Venki: It's a really easy, simple, fast process so that the data and IT teams can react to business needs to changing business needs faster than before with the flexibility. That we can provide, and of course, the result of all of this is a larger business impact. So this becomes directly tied to different business outcomes that we are talking about.
Venki: Chris, I see some questions coming in through the chat. So if there are
Chris: just one question, can you please elaborate a little bit on identity resolution approach here? How is the data stitched together with transactions, for example?
Venki: Absolutely. Yeah, I will touch upon that a little bit more in detail when we get to our flexible entity resolution network models.
Venki: But the simple answer to that is, as you bring interaction data into LTO we have the ability to associate that automatically with the customer [00:17:00] profile. Based on the identifier provided by a source system. So let's take an example of a support interaction that is coming in from a system like a Zen desk or a sales force service cloud.
Venki: There is a customer identifier already present in that interaction, and then based on that, we can associate that interaction with the right customer that is managed inside. But what we will also do going forward is, we are also looking at enhancing capabilities of extracting or recognizing specific entities out of an interaction or a transaction.
Venki: So in, let's say, in payment processing, you're getting transactions that are essentially accounts payable, invoice transactions, that transaction has customer name, customer address, or other identifying information. Can we use that to extract the customer information out of it? And match it to the existing information in rel to you.
Venki: So that kind of capabilities are also something that we are planning. To enhance inside Reltio, but it's really a straightforward process today where when you submit, when you [00:18:00] add or create an interaction in Reltio, we will automatically associate that with the unified customer view because the same crosswalk ID for more, for those of you who are more technical based on the same crosswalk ID, we will associate the interaction to the customer.
Venki: Thank you. So we were talking about the. business benefits and how it accelerates business benefits. And I want to touch upon this one really important aspect, right? All of the investments that we make in our data management data governance, master data management kind of capabilities, which leads to, customer 360 here that you're talking about is and has to be in service of specific business outcomes.
Venki: I don't think there is any dispute to that, right? Which means right from the get go, right from when we plan such initiatives, we have to understand what are the typical business initiatives and what are the categories of those initiatives that we can influence positively [00:19:00] and you can see on the extreme right of the slide, the typical investments are made to accelerate growth, improve operational efficiency or to manage risk and compliance.
Venki: And there are several key aspects and we have created this framework of, there's nine boxes and you can change this and add your own specific initiatives. But if you look at these initiatives across growth, efficiency, risk and compliance, they all rely on some core data foundation.
Venki: And most of them in this particular case, it's all about having that 360 view of the customer. So when we think of customer 360 data product. We make sure when we, as relative, when we look at it, we think of how do we create a product and a set of capabilities in including a data model that powers it, that can then map to these different business outcomes that you see on this, and I would also encourage every one of you, as you think of implementing this, or even planning out these [00:20:00] initiatives in your own companies to really start with those business initiatives, The current state, the future state, and how having this unified view of customer is going to impact that positively.
Venki: And that is an important piece because that helps us prioritize the right areas, the right kind of investments. And it also helps us clearly articulate the value of these investments that we are making in customer 360. So that's how we think about it. So when we think about the relative customer 360 data product, we have that multi domain view of the data with customer at its center.
Venki: Okay. But with information related to it, like households or the relationship between an individual and the organization that they are working with products that are related to these individuals and other related information. And all of that is in the service of improving one of these outcomes, like omni channel engagement, customer centricity, sales effectiveness, or privacy and consent management, right?
Venki: These are some of the examples. [00:21:00] And then we think of, how do we bring the data from multiple different sources, unify that into this data product and then make it consumable through these different channels like digital applications through real time APIs, advanced AI, machine learning, analytics and data lakes so that they can then be used to connect to these different outcomes on the right side.
Venki: As you can see,
Venki: let's get into some more details. Let's talk about what specific functionality are we talking about? Because I, I talked at length about, the need for customer 360 the required, the, the need for tying it directly to business outcomes and how we can do that and how we have as well to your part about that as part of, building our product, our customer 360 product, when we talk about key capabilities, the kind of information that we bring together basically includes first party data sources.
Venki: So any information that is being collected, captured by your systems directly about customer. That is an important aspect. The second part, second [00:22:00] type of data that we bring together then are the third party enrichment sources with which we can enrich the data. So think of capabilities like data providers like axiom or live ramp or done in Bradstreet or zoom info.
Venki: All of them provide critical pieces of information about organizations and people that you work with your customers. And that is another important aspect of the customer 360, right? When you think of information that you want to combine, assemble together into that unified view, first party sources, as well as third party sources are important.
Venki: The third category of information or information about interactions or transactions, right? What specific engagements are these customers having and how are you capturing those interactions or transactions that are being executed by that customer, because that information again, is critical to When you look at that 360 view of that customer.
Venki: If I take an example from an insurance industry, when a customer calls, they're calling about either a claim or a or a, an [00:23:00] existing policy, but maybe related to some changes to the policy or some such thing, or probably to buy a new product, those things, if they are captured as interactions, and there is a single view when that contact center agent picks up the call or connects to the customer through a chat or whatever channel of choice.
Venki: They will have that relevant information as part of the 360 view, which means they can provide more contextual experiences to that customer, right? So that is a third category of information that needs to be combined, brought together to, into that, into the customer 360. The fourth type of information that we talk about, talked about, talk about are derived attributes.
Venki: Derived attributes are information that are derived from existing data, and that could be calculations that you do. For example, total order value of all the orders a customer is customer has had with a company. That's a simple calculation of adding up all of the orders or all of the orders within a specific time frame, but it could also be predicted insights like a customer [00:24:00] churn propensity or a customer health score or specific, next best action recommendations based on all of the different information that a customer that we know about a customer.
Venki: So all of those kinds of information are what we call derived attributes. These are set of attributes or information that are derived. Based on all the information that we have and the simple way of thinking about it are the examples that I mentioned, right? Simple calculations, aggregations of information or predicted information or derived, any information that is created by, let's say a machine learning model that predicts a certain category or certain outcome, or even things like next best action recommendation.
Venki: Those are all classes of data or information that we would call derived attributes. So if you have a 360 that combines quickly, sorry,
Chris: Can you please describe the gin AI UX stuff?
Venki: [00:25:00] Yeah, i'm coming there. Okay. I think people are reading ahead in the slides, i'm just getting Perfect Yeah. So we talked about the type of information now, what we are doing on top of this.
Venki: So in addition to enabling companies to combine all of this information to create a 360, what we are doing is obviously, one of the superpowers that, that relative has that allows us to combine all of this as the connected graph, which is a foundation for data model essentially is an entity graph model.
Venki: So it can combine all the different types of information related to a customer, and it can also link interactions and transactions to that. So that is foundational. The second part of what we're talking about is. The identity resolution or the entity resolution process itself. We've augmented that with applications of AI specifically using large language models for comparisons of different entities or different representation of a customer.
Venki: And that is what we call FERN or flexible entity resolution network. [00:26:00] And I have a slide where I can go into some more details of FERN, but that is included in our customer 360 product, right? When you get our customer 360, it comes with the FERN models that are powering an augmented entity resolution where you don't need to configure a traditional style match rules and then go through an expensive process of tuning those.
Venki: And getting to the right results, you're out of the box, getting a lot of that entity resolution capability powered by AI. And the third aspect is providing a simple conversational experience on top of this data. And that is extremely important because imagine you've got this powerhouse of data, your customer 360 that has all of this information.
Venki: Ultimately, the value of that comes from driving consumption. So you can think of on the right side of this. There are different consumption scenarios, right? You're pushing the data down into your A. I. Machine learning models or gen A. I. Models. You're pushing that information into applications or reporting [00:27:00] capabilities or B.
Venki: I. But there are many situations where I just have a question about a data or I just need to know something about a specific profile. And I want to get to that information as quickly as possible. Traditional ways of doing this would involve going through searches, going and looking at a profile and then trying to understand or make sense of all the information that you're seeing on that screen.
Venki: But what if you had a simpler conversational experience that which allows you to ask a question like. How many customers do I have in the state of New York? Such a simple question, but you can think of how you can add more complexity and say that how many doctors do I have in the state of New York with a specialty of oncology?
Venki: Or how many customers do I have in the state of New York who have purchased a product from us with an order value or net order value of over 1, 000 in the last 30 days. Now you're getting to more complex segmentation scenarios, and we are enabling all of that through a gen AI powered user experience, a conversational user experience that is [00:28:00] powered through through generative AI.
Venki: And we are able to provide simple responses to those questions. So those are the foundational capabilities we are combining in the context of our customer 360. So just a quick recap, it is all of the customer data that includes first party data sources, third party sources for data enrichment, interactions and transactions, derived attributes that are provided through AI powered predictions or other machine learning and data science models.
Venki: Thank you. And then on that data, we enable faster resolution of information, faster identity resolution or entity resolution through our phone capability, which is a flexible entity resolution network. We leverage the LTO connected graph. For managing an entity graph of all the information so that the interconnected nature comes into play when you're looking at a 2360 [00:29:00] degree view of your customer data.
Venki: And the last part is we are enabling simpler consumption capabilities through a conversational interface that is built on top of this that allows for advanced segmentation with very simple natural language prompts and queries. So if you have all of this information, what are the benefits that it might provide?
Venki: And here is an example use case of customer service. I won't go through that again, but we talked about a customer service use case in the context of a I have an insurance company just before, but you can imagine how you can extend this from customer service to a personalized experience on a digital channel to a sales experience or any other part of your organization, which interfaces with the customer directly.
Venki: So what did, what benefits does it provide for the data team? The data team, now you're able to support all of those different initiatives with rich, accurate profile. And with a higher efficiency with a power data management capabilities that provides more automation. So [00:30:00] lower effort, higher output and the ability to keep up with the business demands, which means you can constantly iterate on the customer 360 product that you're creating to add additional capabilities, additional data attributes, additional derivations so that you're incrementally delivering more and more value to your business.
Venki: We touched upon two key capabilities. I just wanted to once again just summarize those for you. So first one is the conversational experience or relative intelligent assistant for data products. Again, just a quick reminder. We did a community show on relative intelligent assistant.
Venki: So if you go into our community and just look through our events you will see a recording of this particular session that goes into it in much more detail. But the high level summary is how can we reduce training time and drive higher adoption by providing a simpler experience? How can we provide more dynamic and rich analysis of the data [00:31:00] that you're managing inside your customer?
Venki: 360 raise data quality and increase data to a productivity all at the same time using these conversational experience. That is what we're enabling with Ria, and I'm happy to show you a brief demo of that in a few minutes, but this is an exciting new capability that we have added to a product. And we have, um, plans to continue to enhance it with adding more and more skills into real,
Venki: the next important capability that we talked about was how we are applying AI. For faster resolution of duplicates or what is called as identity resolution in the context of customer or entity resolution more broadly. And this is where we are combining our large language models that are deployed and managed inside relative perimeter, but security as an important consideration.
Venki: To provide you with the ability to compare different data and get the results, get to resolving those [00:32:00] duplicates in a much shorter timeframe with less effort. The advantage that we see with large language models is these LLMs are by nature built for string comparisons and they are able to handle a lot of sparse information or information that is written or described in different forms.
Venki: And you can see some examples here, right? When there are two different product descriptions for you to have a a traditional sort of comparison that can identify these two strings as same or similar or what level of similarity exists. It's not a, it's not a trivial thing, but with an LLM, it can compare these two representations and compare with its rich source of information.
Venki: And we'll say that, there is a 87 percent probability. These are referring to the same product, but just described differently. That same thing applies to other types of data like names LLMs have a built in understanding of names in context, for example, gender, specific names, female, male names, or nicknames where it can [00:33:00] identify that, Steve is most likely the same as Stephen or Rob is most likely same as Robert and those kind of things without having to train or without having to create a dictionary of such things.
Venki: LLMs can also compare names across languages, not just names, any kind of strings across languages. They have the inherent capability. So now you can compare different representation of a name across different languages and identify whether there are matches or not. So all of these are amazing capabilities that are available to us now.
Venki: With which we can now augment our ability to match and merge information. And that is exactly what we have done. So we've created a network of such large language models that can be stitched together to look at an entity in its whole to look at a customer, for example, in its whole with name identifiers.
Venki: Addresses and other types of information and can compare and with a high degree of confidence can say that the two customer representations, two different records are the same or not. So this comes with its zero short learning, which means there is no [00:34:00] rule creation. As soon as you deploy phone, you are starting to see those results.
Venki: And I'll again show you an example of this quickly here today. It will identify new types of matches that a rule will find really hard to detect to identify. It comes with data security and privacy built into it, which basically means no data. None of your data is being sent out of your tenant context.
Venki: You're not leveraging a public LLM. So security is automatically taken care of explainability. Why two records actually match that is actually built into this capability as well. And of course we can support more than 60 different languages. Natively because these LLMs are capable of of handling that.
Venki: So this is another key functionality that we have built into our customer 360 product that helps with the faster time to value.
Venki: So we talked about some of those examples. Why don't we see what it looks like? But before I jump into the demo, Chris, can you check if there are any other questions that I need to [00:35:00] address right now?
Chris: There's several questions. Maybe quickly, did you say Fern was gen AI driven?
Venki: Yes. It's technically speaking, it runs, it uses LLMs behind the scenes and it stitches them together.
Chris: Got it. I think we have several questions, but let's do the demo and then I'll start asking all the questions.
Venki: Okay. Okay. So first thing you're looking at a customer 360 tenant here. I just want to point out this particular pop up that just showed up. We have, and Chris mentioned this earlier, we have our 2024.
Venki: 2 release coming up very soon in June and mid June. And next Thursday, we have a customer and partner webinar on the 22nd of May. And you can see that pop up inside your. Your tenant, if you log in today or is already published on our community events calendar. So please go register. Love to see you all there.
Venki: A lot of important functionality that we are bringing forth to help you get more value out of Reltio. [00:36:00] So we are excited with that. So I just wanted to Provide a quick plug for that. And so let's just make sure that I'm not logged out of the system. What you're seeing right now, as soon as I go into this is a full page experience with this conversational assistant, right?
Venki: So as soon as you come in, you're not looking at, charts and graphs and other things. The system is basically prompting you to ask it questions like, what do you want to know? You can start with something as simple as, how can we help me? And obviously it will provide you some basic prompts.
Venki: Or you can also ask questions about data. So there are two different types of skills that we've enabled through our intelligent assistant. As of now, the first type of skill is really around content. You can ask the, ask questions about concepts, like what is a crosswalk in Reltio or how can I review potential matches in Reltio?
Venki: And those concepts obviously will be will be summarized and will be presented to you. That is one [00:37:00] type of information that we have or one type of skill that we have developed. The other type of skill is all really around data. So you can ask questions about let me just start a new thread here.
Venki: You can ask questions like Find all individuals in this tenant and you're now interacting with the data and we are as a system is now summarizing data from the system. So you can all you can see the results from here, but it's not as simple as, finding all the individuals you can go in and start making your questions more specific.
Venki: And the system will automatically react to that. So let's say find all in the data.
Venki: Let's see if it comes back with results. So I'm progressively making these questions more specific, right? So now it came with answers to all individuals in the state of Arizona. Can I ask a question like, uh, find all[00:38:00]
Venki: so I can now for the filter and what it is doing is it's trying to translate some of these, whatever prompts that I'm asking, all these questions that I'm asking. And it is now trying to specifically change that into a query. And you can see that it was not able to find any data. Most likely it is missing data, but it is able to change the question to addresses in Arizona.
Venki: I didn't say anything about address. I simply said state of Arizona. And it took this managers and it understood that most likely that is related to the title of salutation as manager and was trying to look for that information and did not find anything right. So the system is intelligent enough to actually look at these different prompts, translate that into specific queries.
Venki: And if I go into this views on the search page, you can see that whatever question I asked, ultimately got translated into a query without you having to know anything about the data model or the specific name of the field or which field you're even searching on. You are just describing a specific type of [00:39:00] information that you want to see.
Venki: And Rhea was able to translate that information into a structured query and will bring up the results. So now what we are doing is we're trying to add more skills that can get into more complex scenarios with respect to querying data. So this becomes a powerful experience. The other thing that we have also heard around RIA from customers is it's great that this experience exists.
Venki: Can I now take this out of Veltio and make it available through other channels? Can I have this executed directly from my CRM? For example, if I have questions about it, or can I give this as a standalone agent? Or a data exploration capability to anybody who needs that information. So that concept of plug ability, as in taking RIA and making it available outside or LTO in different places is definitely part of our roadmap.
Venki: We are looking at how do we solve that? The other aspect is maybe even bringing additional models into RIA so that it can augment the customer information or any kind of information that you're managing with external sort of [00:40:00] models. So there are multiple different things that and directions that we're exploring to further enhance this capability.
Venki: And this is really exciting that we have a starting point, an interface that understands natural language prompts or questions and is able to translate that into what specific actions that needs to be taken in our system in customer 3 60.
Venki: The second aspect that I wanted to show, we talked a little bit about phone and and all the different types of data that is being managed. So here is an example of a customer 360 profile that is managed in Reltio. And as you can see, a lot of this is familiar to you. Even when you're using MDM capabilities, you're looking at creating a single authoritative view of your customer data.
Venki: So there are all the different types of information, all the different attributes, relationships we talked about some of the specific outcomes that you could drive. So privacy and consent related data is being, is captured here. And you can see all of this information. Some of this might also be augmented information coming from third party [00:41:00] sources, right?
Venki: So that is another important aspect. But you will see all of that. And if you were to go into our sources view, you should be able to see the different systems that are contributing data and what kind of data is being brought together. But in addition to that, we also have interactions. So when I look at this particular profile, I can see every single interaction.
Venki: And here I have a number of web events which are related to this customer, but I'm capturing some specific information from from the website in the, in context of this particular customer, what we are planning to do is to further enhance this interaction UI itself to provide a customer journey view.
Venki: So you can look at a timeline and look at the different events that has happened across multiple channels. So that provides more context when you're trying to understand what journey a customer has taken before he or she is interacting with with an application or with a person, right? So these are all the base set of data capabilities that we already talked about.
Venki: We also talked a little bit about flexible entity resolution networks of phone and how it can provide [00:42:00] higher quality, higher accuracy matches. And you can see the results of that right here, right? So what you're seeing under this. Moniker match IQ right now is actually powered by our flexible entity resolution model for, the phone model for individuals where it is looking at different attributes, it's comparing those things and basically saying that there is an 85 percent probability that this particular profile guy Harper is the same as the profile that we are looking at.
Venki: And you can then look through and see what is exactly matching. What is different between them as not only providing that one match is providing different matches, right? In this case, it's saying it is a 77 percent match as opposed to the first one having an 85 percent match and what this allows a data steward to do.
Venki: If you have a steward acting on this information. is to more confidently go ahead and take actions of merging this information or not merging. But where it becomes really interesting is where you don't even have a data steward, [00:43:00] where you can set thresholds in the system. And the way you would configure that is in our data model under match.
Venki: Now you have the ability to activate this match IQ or the phone model for individual. And if I were to go in here, I can set thresholds, but I can say that any match that is identified to be over. 95%. I can in this particular case, it's recommended for merge, but I can ask simply change this to automatically merge, right?
Venki: So a lot of those high confidence matches can be not just identified by the system. But can be automatically merged into a single consolidated view by by the system. So these are some enhancements, some critical enhancements that we are adding or we have added to the product in the context of customer 360.
Venki: And we will continue to enrich this product as we go forward so that we can provide a high quality consolidated, 360 degree view of your customers. We can enable consumption of [00:44:00] all of that information in real time or in near real time channels. We can augment that information with machine learning based predictions, and we can enable enhanced segmentation capabilities and data exploration capabilities through this conversational experience via the relative intelligent assistant that we have created.
Venki: So that was a very short version of a demo. It was not meant to be an end to end demonstration, but I just wanted to give you some visually just wanted to show you what those concepts and what those what those capabilities mean. So hopefully that helps
Venki: us. Great. Moving on. Let's talk about what is coming next, right? There is, there's more, there are more exciting things. So there are planned enhancements. Like I said, we are constantly adding more and innovating. So we're talking about how do we enhance conversational experiences with more advanced skills where I can even go into a profile and I can ask for a summary of the profile or highlighting the most important aspects that I need to know [00:45:00] about a profile.
Venki: There are more advanced segmentation capabilities itself, which includes. Interaction data. So I can segment based on certain things like all customers who have made a purchase in the last seven days is a segment that I would like to know. So how can I pull those kind of information together? Integrated AI driven predictions.
Venki: So how can I include a customer propensity model into the data in customer 360, where you don't have to go build that separately, but just go and activate a model and have the predictions available directly inside. Do you So there are a number of those capabilities around AI predictions that we are looking at, including looking at some of the model gardens that are being created in the cloud hyperscaler ecosystems, right?
Venki: Google or AWS, they're all coming out with an ecosystem with a library of models that can be consumed through SageMaker or through Vertex AI and others. So how do we integrate some of that into Reltio's Customer 360 is another aspect that we are looking at. We talked about [00:46:00] enhanced segmentation and activation capabilities.
Venki: I will actually show you a few screenshots of what is coming in the upcoming release in our 24. 2 release. We also continue to enhance our data pipeline so that the data can be consumed easily into cloud data warehouses. This is in addition to all of the integration capabilities that we provide today already.
Venki: That helps you send your data from Reltio into any of the consuming systems. And through relative integration hub, we have already, about 1400 plus connectors to all the major marketing automation systems, CRM systems ERPs and others where data gets consumed. So we are also adding to that consumption to cloud data warehouses natively.
Venki: And as of today, we have Databricks, Google BigQuery and Snowflake supported. And we're looking at adding additional ones like AWS Redshift. As your one lake and others, and then we also are integrating to data catalog solution. Now, the concept of a data product, one of the key [00:47:00] concept that it emphasizes on is self discoverability of data, which means people in your organization.
Venki: How can they identify? What is the, what are the data elements in a data product and other aspects around it? So that helps with consumption. To enable those kind of scenarios, we are also integrating with data catalog. And the first one of that is with Colibra that is coming out fairly soon in our upcoming release.
Venki: So to know more about those things, please do attend our release webinar. Here's what our segmentation experience is going to look like in the upcoming release. So now there is going to be a dedicated place where you can create a segment out of the different characteristics of customer data that you're managing.
Venki: You can go into that into the segmentation view. You can identify different criteria based on which you want to do the segmentation. You can then save that specific segment, and you can also share those segments with other people so that the segments that you're creating is visible and be consumed through other channels.
Venki: And we are integrating some activation capabilities where we will [00:48:00] link these segments to specific activation flows or recipes that are created using relative integration hub to push this information into specific downstream systems. So if you want to take, let's say, highly engaged customers and push this out into.
Venki: Spot and execute a marketing campaign. You can define a recipe through integration hub, and then you can link that into the segment and push it out as an activation.
Venki: So these are, this is just a glimpse of one such capability that we are adding to the product. There is more to come. And like I said, that we will be sharing more details about this in the upcoming release webinar. So let's try and wrap up. I know we are coming to an end. We've been talking for almost two hours.
Venki: For 50 minutes now. So very quickly what relative customer 360 data product enables companies to do is to create comprehensive profiles of their customers with relationships, [00:49:00] interactions, derived attributes. It is powered by phone for entity resolution for faster, higher quality interview resolution.
Venki: It comes built with a gen AI powered conversational user experience. for data exploration. And obviously what is built into it, the multi domain capabilities, which means you're not just looking at customer data as an organizations and people, but you're looking at related information that can be added to this for a full an augmented view of that customer like subscriptions, products contracts locations, or stores.
Venki: Programs or other kind of things that you want to track in the context of a loyalty management system. Any one of those things can be modeled as part of our extensible customer 360 product to bring those multi domain capabilities together and all of this information can then be used to power the different business initiatives to accelerate growth, to improve efficiency and to manage your risk and improve compliance.[00:50:00]
Venki: So what makes relative customer 360 product different? It's enterprise wide source for enterprise wide use cases. It's not siloed. It's not. It's not limiting to a specific department or a departmental set of use cases. It comes with a semantic understanding. We have that data model. And we will also have data model or customer 360 customer 360 versions for different industry segments.
Venki: But today, customers can already leverage our platform capabilities to create a customer 360 that is more suitable for an insurance, which is probably slightly different from what is required for banking, which may be different from what is different, required for high tech, but our customer 360 data product supports all of those different industry scenarios through the configurability that we support.
Venki: And we built this on a data as a product approach, which means all of the capabilities that you need to understand your [00:51:00] sources, your consumers, quality of the data consumption and performance of consumption, all of those are built into this. So the product owners for this can truly enable different business outcomes and different use cases with treating this customer 360 as a data product.
Venki: Okay. And obviously this is built on the same platform. So all of the relative platform differentiators come along with that, which includes real time secure and interoperable data. Highly flexible and scalable architecture. AI is built into different aspects of this product. As you already saw through our conversational experience through phone or through those derived attributes that we talked about in the different AI models we are bringing together in the context of the customer 360.
Venki: And all of this essentially is to drive faster time to value with prebuilt configurations and components.
Venki: So I think that was most of the [00:52:00] content that I really wanted to talk about. This is the last slide I have for today. I really hope that you got a good end to end view of what customer 360 data product offers, some of the key capabilities that it provides and how it enables faster time to value by enabling companies to create a ready to consume 360 profile in a trusted and secure manner for addressing any business use cases for delivering value, delivering real tangible value for businesses.
Venki: And it is not a static product as you can see from the roadmap or some of the planned enhancements that I shared, we are investing heavily. We're adding more and more capabilities and enhancing some existing capabilities very fast. Next week, you will hear some more about some of those things.
Venki: Thank you. But over the next few months, you will see how we are at, enhancing our customer 360 to be a high value product for any customer who's thinking of investing and [00:53:00] creating a 360 degree view of their customers to power their business initiatives. So I'll stop there because of the other questions.
Venki: I'm happy to take that right now.
Chris: Yep. Let's get with it as many as we can. So what is the back end database for customer 360? And what tools are there if we want to ingest huge volumes of data and customer 360?
Venki: That's a good question. So customer 360 is built on the relative connected data platform, which basically means it runs on all three clouds on AWS, GCP and on Azure.
Venki: On each of the clouds, we have an optimized stack that leverages some of the managed databases across those. So the database, it's not a, it's not a single answer. It depends on the cloud of choice, but irrespective of that, what we ensure is scalability of our infrastructure, our services. So you can bring in large volumes of data into Reltio.
Venki: For interaction data right now, we have the ability to consume them or create those [00:54:00] interactions, ingest those interactions through our APIs. We also have a mechanism where you can drop them into a queue and you can use a recipe to essentially ingest them on a more continuous basis. So there are web events, those kind of things, then it can be consumed through that.
Venki: And if there are specific questions about volumes or concerns about volumes we are happy to address that. But. We are already supporting customers who are bringing in several hundreds of millions of interactions into Reltio.
Chris: Thank you generating derived attributes currently requires customization, wherein we need to either derive the attribute and inbound integration or through a custom code like LCA, et cetera, or pull the data out of Reltio to calculate and write it back to Reltio.
Chris: Such as. Reltio360 simplify this where we can derive attributes through, some simple out of the box configuration or how does AI features help deriving these derived attributes?
Venki: Yeah, [00:55:00] again, great question. So what we are doing differently, there are two, two options that are available where we are investing more.
Venki: One is how can we, through the connectivity that we're providing to cloud data warehouses, like a Databricks. How can we provide a more seamless capability for those predictions to be applied against that data in that warehouse and then to be brought in the results of those things to be brought into Reltio.
Venki: So that is one option. Granted, it requires some configuration or customization today. But that will be the recommended approach, which many of our customers are already using today. And the second is we are exploring integrations with certain low code or no code pre built prediction capabilities from some of our partners.
Venki: And with that integration, then you will be able to just enable those integrations with Reltio. And those prediction engines will know exactly how to predict information, how to access information, how to predict those outcomes and how to [00:56:00] make those predictions available as part of those derived attributes inside your LTO data model.
Venki: So that second category of information is still in the works. You expect to hear more from us. As soon as we have it ready, our plan is to have some of those integrations ready to be showcased to you within the next 6 to 8 weeks.
Chris: All right, I'm going to ask one question because I know everybody wants to know and then we'll get to the others later, through the community.
Chris: So can you please quickly touch on what types or levels of Reltio licenses are needed to enable all capabilities for this 360?
Venki: So Reltio Customer 360 is a new product and so you'll need to license that particular product and it comes with all of the capabilities that I talked about. So I think that is probably the short answer.
Venki: I think there is always a long answer when it comes to licensing and I'll let my, the experts in our sales team put questions, but please do reach out. If you are a partner and you have the same questions, please do reach out to a partner or [00:57:00] alliances team, and they'll be able to guide you through that.
Chris: Awesome Venky, that's all we have time for today, everyone. Please take the survey at the end, and if you are interested in looking at what is required for a license and things like that, there is a question in the survey to add your name and email, and we'll make sure to reach out. But thank you, everyone, for coming.
Chris: We'll definitely get these questions answered on But until next week, and we'll see you at the release webinar. Thank you everyone for coming and spending some time with us. Thanks, Venki, so much for this information. Really awesome.