Download the PPT: Redefining the User Experience with Reltio's Intelligent Assistant PPT
The onset of GenAI era is revolutionising the customer experiences. Reltio is at the forefront of this innovative and exciting times in the world of AI and has launched its own GenAI-powered smart assistant -Reltio Intelligent Assistant (RIA) that helps answers any queries about its data assets, anytime and anywhere.
Transcript:
Chris: [00:00:00] All right, so welcome to another Reltio community show really excited about this one today, on redefining The user experience with Reltio's Intelligent Assistant or aka RIA. We have our special guest Vidhi Chugh. She's the Director of Product Management of AI and ML. Welcome, Vidhi. How are you?
Vidhi: Hi, Chris. I'm good. How are you?
Chris: Doing well, and this is your first time, so I'm really excited to have you on. She's our AI and ML guru that I've been super excited to get her on a show, and she'll be on more shows here in the future so the rules of the show are keep yourself on mute, feel free to ask questions in the chat we definitely have some people on that could help answer questions, not going to be a real technical show today.
Chris: But we will record this and we are recording it and it will be shared out probably tomorrow [00:01:00] once I get that edited and cut. And then we have some upcoming events coming up just on show wise. I'm not going to go through all of those, but today's show on RIA. We do have one next week that I'm super excited about as well on efficient, empowered, engaged, leveraging Reltio UI for data steward success.
Chris: A lot of really good stuff coming around that here in the not so far future. And then several others. So you can take a look at that. And then if you're a life science customer and going to be in New Jersey on May 9th, we do have an event coming up. I'll put that link here in the chat here shortly, but it's a free event.
Chris: Register now. It's a half day. We'll have Manish, our CEO, as our main speaker, and then we'll have a couple of customers talking about their journey with RelTL and MDM. And then, is that all my slides here? Yeah, [00:02:00] that is all my slides. So that was an abrupt ending. But Vidi, I'm gonna hand it
Vidhi: over to you.
Vidhi: Yeah, thanks, Tess. Hi, everyone. I'm glad to have you all here. Join me. I am going by introduction. I'm Vidhi Chugh. Chris already introduced me a bit. I'm heading the product, the AIML side of the product. And yeah, I think I'll share the screen. Just give me a moment.
Vidhi: Perfect. Okay.
Vidhi: All right. Can you just confirm my screen is visible? I can.
Chris: Yes. Looks good.
Vidhi: Okay. So hi everyone. Welcome again. I'm pretty excited to share all those AI ML innovations that we've been doing at Retio. Today's webinar is going to be on one such AI ML offerings. And before I start with that, the topic [00:03:00] today is on redefining the user experience and I'll.
Vidhi: In the coming slides share how our very own Reduce Intelligent Assistant is redefining the customer experience. And I know that I've spoken about AI ML innovation and I'll just double click on that before I proceed further. Generative AI, I think has revolutionized the way the businesses operate today, be it in terms of, opening revenue streams or controlling cost or Having defined processes, which can help you gain more efficiencies, right?
Vidhi: So in terms of all those generative AI related offerings that the enterprises understand, they also know that all of those data assets That they have are their gateway to unlock that time to value the basis on which they can take business decisions. And that's precisely what we are also ensuring by making those innovations, which are generative AI powered a [00:04:00] preview of which is what I'm going to discuss in today's webinar as well.
Vidhi: How we are pioneering in that space by leveraging generative AI for unifying and managing data. And before I go further and, talk a little bit deeper about original agenda, I'll just quickly give a preview of our mission and, share how we are going to do that. And we understand that because data is the asset that the organizations are looking to tap into its potential.
Vidhi: We also understand that. They need to have a trusted, unified and interoperable data view based on which they can act on the output coming from that, those data assets, which could be used in terms of improving the efficiencies, manage any kind of risk thereby enabling the business growth. This is our mission.
Vidhi: This is what Reltio's mission is. And in order to do that, we are making a lot of investments in this space. One of our key themes for 2024 is to ensure that [00:05:00] customers are empowered. Our customers are empowered in terms of how they want to access the information. They are, it's more about self serve. They are able to help themselves and there is, we are ensuring that there is a path of least resistance.
Vidhi: If they're looking for some information, they don't need to rely on any kind of. expert who comes up with domain knowledge and then I am going to ask them some information and by the time I get the response it's considering all of these things, it's an impediment for me to get access to information, right?
Vidhi: And in today's world, the faster I enable that path, The faster I remove those impediments and make them have that access to information is the, is directly proportional to the time that they're taking to unlock the value that comes from those data assets. This is what I call as unlocking that 10X productivity.
Vidhi: This is our goal. This is one of our key focus for 2024. And in order to do that. [00:06:00] Is the part where our innovation lies. We are working on religious intelligent assistant. That's called as RIA that we discussed in the beginning. That's the place where RIA enters and what it does is, and I think you would be asking that, the name itself says it's intelligent assistant, and we are otherwise also seeing a lot of smart assistants.
Vidhi: I think the beauty to, to appreciate the beauty or how these assistants are making our lives easier. I think it, it's important for all of us to also understand how these. Assistants were, how we were interacting with assistants when they were not intelligent. And we, I am sure that we all must have had experience before when we were using assistants and chatbots which were just used for search purposes, right?
Vidhi: They were not able to gather us insights. Or be able to understand the context, right? If there is a contextual information, those assistants were not able to understand which is what we are now able to do. That's what Reltio's intelligent assistant [00:07:00] Riya is able to do, right? So if it's good to see the evolution, how we are, coming from those previous traditional ways of thinking of assistants and how we are redefining our customer experience with the launch of Reltio's intelligent assistant.
Vidhi: Now, when I talk about RIA, one thing that specifically clicks to me, and I really want to share with all of you is it ensures data towards productivity and how. Let's compare how without this RIA, the data stewards were looking for information, which used to require manual effort in terms of searching for that information.
Vidhi: For example, if I'm a new person and I don't know about this domain or any question that I can have, and I need to look for a response to that because there is some action I need to take on top of that information. I would add either go to a person who is an expert into that domain and say questions like that.
Vidhi: I'm showing on the right side that how do I load data into Reto? I know I could [00:08:00] have question about match tool configuration, or I would like to know about how to delete a contributor, right? These are the questions. If I have these questions and I'm looking for the information, either I have to look for a domain expert or.
Vidhi: I can search around with the Red Tools digital assets and then, comes a lot of links. I need to open them, understand the content into those documentations, and then make sense of the information and curate it for myself before I act on it. You must have seen that by the time I'm explaining this, there's already, we can sense there's a lot of cognitive load that the data steward has to go through in the absence of RIA.
Vidhi: Now, when we know that RIA is not there, And this is our conventional way of looking for information. Now see the kind of advantages that comes with RIA. It's amplifying the impact by eliminating those several human hours that were otherwise spent in looking for information, that was covering for information, which is now not happening anymore.
Vidhi: If I do not have access to a domain expert, what I would do otherwise is look for [00:09:00] A way to raise a support ticket and then wait for a response, right? Until that time, I'm blocked. I'm not able to proceed in any of those initiators that I was originally working on. All of these issues are now handled by a generative AI based powered assistant, which understands the language.
Vidhi: in the similar way that we humans talk, right? It's, that's why it's called as natural language, right? So that's one thing that clicks with me definitely. And I keep on using VIA for all questions that I have, and I encourage all of you to use that and make use of it and see the kind of responses that it gets approved a glimpse of his of which is what I'm going to show in the demo as well towards the later half of our webinar as well.
Vidhi: But if I go further, I'll just say that by the time I explain this, it already looks so easy getting the access to information with RIA that The time to access to information that I mentioned is directly related to the time to value. That's what we are seeing here that it immediately speeds up my time to get value.
Vidhi: I have information on fingertip anytime, anywhere [00:10:00] with an assistant, which is next to me available 24 seven. Right? So 24 seven, it's available. I can ask question. There is no predefined way I can simply ask it like I'm going to have a conversation with humans. Similarly, I can ask this assistant. It can cater to any kind of depth of questions, any complex technical documentation, any kind of those queries we can ask RIA and it can answer those, right?
Vidhi: One thing that happens with typically all AIML models, and that's true with RIA as well, is That it's continuously learning. It's understanding how it can better respond to your queries, right? And there is a framework that we have developed in house. And I'll just share a bit about this in a little later as well, but I'll give a background on how we are thinking about making VR more useful for the users.
Vidhi: It's continuously improving. It's accuracy is improving. And on the feedback part, I'll just, double click on feedback as well. It's twofold. One way of seeing that, how do I, as a developer of RIA when I'm offering it, know that it's [00:11:00] being useful to the user, data stewards are finding it useful.
Vidhi: So for that we are giving, we are building a feedback mechanism as well, which would be a way to show the quality of the response from the data stewards side by taking on thumbs up if it is of good value for their question, the original query that they asked. Or it could be thumbs down. This is one way to gather feedback from the data store, from the users of VR to see how value, how it is providing value to them.
Vidhi: Now, the second part of feedback is something that I really think is going to be useful and of immense value to the data stores is when we talk as users, we keep giving each other as human, the feedback, right? If I'm talking and I'm asking a question. And for example, there is another human I'm talking to another domain expert, and the intent is not right clear from the very first instance, and I would say no, this is not what I meant, and I meant this, can you give me an, and I start giving an additional context, right?
Vidhi: So that's what if in the form of [00:12:00] prompt in the form of query, if the user is writing that this is not what I was asking for, but I'm asking for this and gives an additional context. RIA is able to, RIA should be able to understand that. That's the feature that's going to come out soon. That's something that I'll be talking about what's coming up next in RIA.
Vidhi: And maybe with Chris, we can have another webinar, which would talk about the entire roadmap of RIA as well. But. This is something I thought would be very useful that if you are you are able to talk to RIA as free as possible, give it those feedback based on the preceding content that you are giving before the context, before giving the context, it can sense that this is something that I need to put more focus on, more attention on this is coming as a way of feedback for me.
Vidhi: So it's a way of saying that it's an analysis of the context that it is giving. It's able to understand the intent and act on it. So that's the part on feedback that's written here. It's twofold. One is on the thumbs up, thumbs down. Another is to try to understand the context. [00:13:00] in the form of feedback that's coming in the prompt from the user.
Vidhi: And this is going to be the version that's coming out soon. Now, the last part of it is when I say that data students were previously spending time on looking for information. And I have a way to understand that finding information manually was time expensive exercise. What happens with Reltio, all of a sudden you have that in time free.
Vidhi: So which was otherwise spent in scouting for information. Now you have that free time at your disposal, which as a data steward can be used for core data stewardship tasks. So now it's can be used for improving data quality and all those four four tasks. And this is right now I'm talking about documentation search and we are also expanding it, which I'll share in the next version as well, that will include community based and, um, yeah, so KB and community knowledge based and community both in the coming releases and it's continuously evolving.
Vidhi: We are going to have new additions on RIA as we speak. So that's about this. [00:14:00] And Chris, I see some questions coming up. Should I take those towards the end? Yeah, I'll ask
them.
Chris: Don't worry. There's not really much but there's a couple, but yeah we'll get to those questions here shortly.
Chris: Don't
Vidhi: worry. Okay, so now that I've mentioned about the advantages of RIA and those features, there is something very exciting coming up with RIA in terms of data products, which is, which I'll say, which is the very core part, which I'll say for the last part of my discussion today, but what if I am not just looking for documentation, I want to extend the way I interact with RIA, and ask questions in the form of, find me all the individuals, right?
Vidhi: Any information about understanding the profile better, or the statistics about those profile, or there are certain metrics that I want to understand it. It's similar way that I would have otherwise queried, right? Now, in the form of natural language, I can query RIA and get those questions, the ones that I'm highlighting here towards in the right side as well.
Vidhi: It's about understanding the profile. Maybe I want to understand the distribution. I'll say that [00:15:00] maybe create a bar chart for me or pie chart or put a filter and tell me what are the customers who are of high purchase value for me. All those questions also RIA would be able to answer. These are the advanced features that we are building, which is again, generative AI based.
Vidhi: It's a chat UI right now, and then you can use it to access and understand and manage your data well. All in all, when RIA comes with data products, the advanced features of RIA is going to give you a dynamic understanding and a much richer analysis and insights coming up from your data. You can ask questions in any form and get those understanding based on which you can take business actions.
Vidhi: Okay, so it's time for demo. I know that I can keep talking about its benefit a lot, but it's good to have a look at the demo first. Yeah, Vidya, it's funny
Chris: you say that because is there going to be a demo? And so we're seeing that in the chat and one person did say and I appreciate that, he's using a couple of times and got answers quickly, in two minutes, rather than taking an hour to research it.
Chris: So [00:16:00] it's been good.
Vidhi: Okay. Should I be taking questions first or?
Chris: No, go ahead and do the demo and then we'll take questions after that. Yep. Okay.
Vidhi: So it's visible, right? Okay, correct. Okay. So the way I call it is on the screen towards the upper right corner, this blue corner. This is what we call as a ML widget. And I call it as wizard. This is where you see all the magic related to here happening. And the moment we open up, this is a side panel. And like we are, like I previously mentioned as well, it's continuously learning and those response quality can vary.
Vidhi: So this comes up with the kind of, a little bit of information about how we can make use of Fria. And in place where it's. Ms Given. As you can ask literally anything to Ria, I'll just start and I'm very polite with the way I interact with assistants as well. So I generally start with the greetings in terms of how Ria, hi Ria, how are you doing?
Vidhi: And then I simply go on, ask questions. For example, I know I need to know about what is the latest [00:17:00] and the release 24.1,
Vidhi: and it's going to gimme a response, which is about. Whatever happened in 24. 1 release, the kind of, and when the preview happened and all those related information, and it's giving a bit about the intelligent assistant itself, the one that we are talking about already. The good part is. The previous one, the previous method that I was explaining is that I need to go through a lot of information or the documentation to get the information, right?
Vidhi: Now it's giving me a summarized view of it. The exact thing that is curated for me to know how to, for me, for the information that is given that is relevant to the question that I originally asked, but what if I'm interested in getting more knowledge, more and more information about that question, right?
Vidhi: Then it comes up with this particular link as well. This link is what is going to help me with the additional information, which if I look in here. I can go and simply check the study four one release notes, which is going to give all that [00:18:00] information in detail, a summary of which, or a curation version of which is what I was seeing in the REA interface.
Vidhi: So I will just go scroll past it first. And this is the one most relevant response for the original question that I had that I had asked Priya. So this is about documentation, but I won't stop here and I'll just keep asking Dia some of the questions. And if I ask it to do some profile searching for me, something that I discussed about the advanced features, I can ask it about find all organizations. For example, from wherever the country is, US,
Vidhi: right? I don't know if I did typo, maybe not, and it's able to do the profile search for me, right? Now I have the list of all of the organizations with country, it's United States, I'll just click one of them to see how it shows up. And I can see the country here, [00:19:00] right? And this is one way to see the profile.
Vidhi: And like I mentioned, if I want to look at the distribution and I say go ahead and create a pie chart for me, right? The one that I was originally talking about, sorry, I think I typed somewhere else. Yeah, maybe create a pie chart of organization by organization type. And there
Vidhi: I go.
Vidhi: This is good. It came up. Okay. So the good part is I can see the distribution. I can see the pie chart by organization type. And the one widget that I see here, which says add, if I click on it, a pop up opens, which says I need to give a name. For example, if I say it's a pie chart. On organization time. And I click add on this.
Vidhi: I can immediately go and see it [00:20:00] being added to the dashboard.
Vidhi: This one is here. So these are the different ways. I was just trying to give an emphasis on how it can do the forming of information through documentation. It's going to be extended to KBN community very soon. I can go and look for information about certain profiles, do those source, put a filter on it and then can see a little bit of visualization because that's an easier way for me to consume the information as well as store and add that into the dashboard, right?
Vidhi: So that's a bit about demo and I'll go back to the some bit of information that I have. Otherwise, it's well planned for the webinar. But should I stay here? Is there any question or I can go and move to the original site? Let's ask,
Chris: let's get some questions. 'cause there's three. So does it support analytic functions like I guess minutes or minimum max aggregator, et cetera?
Vidhi: Does
Chris: it support? [00:21:00] Yeah minimum, maximum aggregator, stuff
Vidhi: like that. Yeah, so I think we are working on this. This is something I need to pass to him. Are you there? Yeah, everything. I'm here. Yeah. Can you take this question? Sure. So currently what we are doing is just fetching the data as well as creating the dashboards on top of it.
Vidhi: We are, we do not have the functionality to take the minima or the maxima of these values. But that seems like a possible way there is a seems like a possible way to do it in
Suchen: the next
Vidhi: sharing this because that's precisely when I gather from a manager as well. This is a quick and easy thing that we can do on top of it and I'll add it to our own map.
Vidhi: So definitely we can do that.
Suchen: Yeah. If I may add something here. Yeah, so as Vidhi demonstrated, there are two areas that Vidhi demonstrated. One is basically searching for content on the [00:22:00] documentation portal, right? If you are new to Relgeo, if you have been spending time with Relgeo for last one year, the time that will take you to find information will vary based on, you know how to search for things.
Suchen: So with what we demonstrated, you can see that, that, that bridge or the gap is narrowed. Now you can just type whatever you want and it searches for the documentation. So that is first area. The second area that we demonstrated was query, right? Anything that you can build from advanced search, from the search query and all of that's the second area.
Suchen: We do not have the, API on the UI today for minimum maximum aggregation, which is why we don't have that as soon as we add those underlying backend APIs, that's when I will learn from it. So the good thing here is the framework is there. It is just about enabling those APIs and we can add much more much faster.
Chris: Yeah, it's just getting right. So we do have an important question here that I think everybody wants to know [00:23:00] about. Will customer data be used to train Ria be shared outside of customer environments?
Vidhi: No. So that's a good question. I think I have covered it in the next part, but I'll take it in a moment now.
Vidhi: So Chris, any question before I move ahead? I'll take that question in this slide. All right.
Chris: Is there, there's a couple more and you tell me if you want to wait till the end, but we'll ask this one. Is there something that needs to be enabled in our environments to allow RIA to access our entity information?
Chris: So when I try to replicate the queries she's performing, I get. A message from Maria that it can only answer questions about documentation. Yeah,
Vidhi: right now it can yeah, the one that we are seeing right now is on the documentation search documentation portal. The advanced features are going to be coming out with data products.
Chris: Got it. So that's in April, is that right?
Vidhi: That's first, yeah, that's
Chris: first week. We have two more questions, but go ahead and then I'll ask the other ones.
Vidhi: Okay.[00:24:00]
Vidhi: Okay. So the way I switch and also mentioned is that the we are able to access the information. The data stewards are able to access the information, which I showed in the demo as well. It is very quick. You just need to put your intent in the form of prompt on a query and you get the response instantly.
Vidhi: So the benefits are multifold. I've just listed very few here because I wanted to, put a very strong focus on this. The first is on relevancy. How do I know if the response is relevant? This is what I'm looking for. And the question that I've asked, I'm getting the relevant response from RIA, because if that's not the case, then again, it will take me to those efforts, right?
Vidhi: And that's what we are precisely saving. So while developing the model, that's what we are doing is to ensure that there's an internal framework that we have developed to ensure that this is the ideal response that should have been this, how typically, and any AML exercise work that you need to have, how the typical ideal response should look like.
Vidhi: And then when we ask RIA the same thing, and the response that it gives us, we try to compare these two. What's the response from [00:25:00] a model versus how the response should have been, which is what we call as ground truth. This is what we are expecting it to give. And that's what we are sensing the relevancy score with.
Vidhi: We are trying to see how far, how near RIA's response is to the actual response that RIA should have given, that a user would have otherwise looked from a human or a domain expert. So that's on the relevancy part. That is the response you're giving to me is accurate, something I can trust on, right?
Vidhi: Because we are eventually using, making use of these offerings. We want to build that trusted view of data. So relevancy and accuracy is one of the core benefits of RIA and like I gave a comparison with the traditional way of having assistants next to us, those that were not aware of context, RIA is aware of context, it knows if in a prompt the user is giving a certain context, if there's also related with Ensuring and keeping a tab on that RIA is not hallucinating.
Vidhi: That's something that typically comes up with the concerns when we talk about start talking about generative AI models, right? Something that I've written [00:26:00] on the right side as well, which is about how do we ensure that all the benefits that are listed on the left come with RIA, but Are there any risks involved or not?
Vidhi: And that's what I wanted to give a very transparent view on how we are ensuring those risks are handled within rel2o and everything that the question that we also asked about data privacy, any of those user concerns we are already addressing before we even, while making and developing RIA. So it's aware of context.
Vidhi: And the third one is about faster time to address query. So I understand that the user has asked a question. What if RIA is. What if there is a scenario where you ask it a question and it gives you a certain response and you give it additional context. And by the time you're iterating with RIA if it is giving you the response right in the first go that's something I call as that without code.
Vidhi: The user taking the extra effort of writing that additional context and putting that effort into explaining what they originally meant, VIA is able to understand that context right in the first go and is able to fast [00:27:00] address that query in the first go itself, right? So that's the faster time to address the user query.
Vidhi: All these benefits are there and these are materializing and it's of high value to the data stewards and saving them. Man, manual hours of search that they were doing previously, we're also ensuring what I'm showing on the right is that we're keeping a tab on the first thing that I've heard in a lot of assistance that, yeah, it gives response, but after a point, I see that the response is redundant, or there's additional information that it is giving, which is not directly related to the original question that the data steward would have asked, or, of any other assistance that the industry is having outside.
Vidhi: Okay. Thank you. So we are ensuring that RIA is not giving any redundant response as well as and keeping a tab on hallucinations that it is not given the liberty to go and fabricate response on its own. It's very factual, relevant and based on the kind of It's a support domain expert, right? So it's not supposed to create response on its own [00:28:00] and we are making sure it doesn't do that.
Vidhi: So it's very on point to the question that the user is asking, as well as we are ensuring that we have those effective guardrails that any question that's the one that I'm talking about in documentation search the documentation portal, this is all nonetheless is the public information available. So all that user asks is the prompt.
Vidhi: Has that intended the question and then just understands that and gives the response. So right now, no user data is being used. No customer data is being used for any kind of model training. So I hope that answers the question that was asked previously, that are we having guidance to make sure that no customer data is being used?
Vidhi: Yes, we are having those in place. Yeah, I think that answers
Chris: the
Vidhi: question. Okay. So I'll take questions. Before I move to the last layer,
Chris: is it available for all tenants or any specific service? We have to add the future.
Vidhi: Yeah. Sushant. Sorry. Is there something you would like [00:29:00] to?
Suchen: Yeah. So Ria, the documentation search and the query search that we demonstrated the documentation article search and topic search that will be included.
Suchen: As part of your base subscription and for the other we are still having those discussions and it's still evolving. We'll get back to you on, on, on some of the other capability, which is like searching, adding graphs and whatnot, right? So those things will be still evolving and we'll add we'll provide more information on how it will be included.
Suchen: In the product.
Chris: Great. Thank you. And does RIA consider tenant specific configurations in response? So example is ask what the match rules in my tenant are. Is this something that will be returned at some point? I can answer
Suchen: that. I talked about the document decision search, right? So that's one area for RIA, which RIA is currently trained on and continues to get trained on.
Suchen: The second part is basically the tenant data. Which you are accessible or which you are [00:30:00] authorized to access, right? So it can access that and provide you the information or help with building the queries at all. The third part is basically the configuration. The rule, is any connector enabled for me?
Suchen: How many attributes do I have? What kind of entity do I have? That's more of a configuration of the tenant itself. So that's a third area. There is something that we are working on. It is not that ARIA is trained on that particular area. You won't be able to, ARIA won't be able to respond to those questions just yet.
Chris: So what type of data quality check can it handle? Is it possible to auto find and correct the profile?
Vidhi: Yeah I'll take that. So on data quality Chris also Anusha I see that question. So we are otherwise as well working on enhancing our data quality initiatives using AI ML right now for RIA. It's on the roadmap. It's planned in tandem with how we are, advancing our data quality initiatives.
Vidhi: So right now the answer would be no.
Chris: Is the [00:31:00] model self hosted or using an internal service?
Vidhi: Imanshu, can you take that question? Is the model self hosted or using an external service? Yeah, so the model is right now GPT powered.
Suchen: Okay. The question, yeah, sorry, go ahead.
Vidhi: Yeah. So are you using Azure, OpenAI,
Suchen: or,
Vidhi: Google's different models now or LLA from meta? What are you using or is it in know, internally hosted? Because we are in a regulated environment, we cannot have external services looking at our data, so we just wanted to understand that very clearly.
Vidhi: Yeah, so data, the tenant data is not going outside, so that, from that perspective, I understand the regulation aspect of it. So that's taken care of. So it's hosted within Realtios? Yes.
Chris: Thanks. [00:32:00] Um, so some clarity on relationships of entities in RIA and APIs. Will all results be from Realtio API calls or some other some, I'm sorry.
Chris: Will all results be from Realtio APIs calls or some other was other providing? So will RIA have any guidance capabilities? So Data areas, potential quality issues and things like that.
Chris: Yeah, I'll post that in the
Vidhi: chat. Yeah. I can't see that question. Yeah. If you can
Chris: come, yeah, it was directly to me, direct
Vidhi: will there, will all results be from radio API call or some other way of. Providing, will you have any guidance capability that is data areas with potential issues?
Suchen: Yeah, if I could just explain a
Chris: little bit.
Vidhi: So 1 thing is the when we're looking at the entity part of Ria. Okay, leave the documentation out any results that are coming back
Suchen: at the end of the day. Are those results
Vidhi: from [00:33:00] the exposed relative APIs? Or is there some other mechanism that it could be providing results from? That's from Relative APIs.
Suchen: Relative APIs,
Vidhi: okay. So it'll be, and that'll be from APIs that in theory we could call directly ourselves if we wanted to. So they're exposed APIs. Yes. Okay. Perfect. Okay. And then the other thing that I can, and then I'll go back on mute is there any kind of guidance capability that's expected to be added in?
Vidhi: The idea of even someone like the data steward type work of maybe trying to identify potential quality areas or even possible ways of
Suchen: improving data.
Vidhi: Just, any plans on that in that direction. Yeah. So I think that's similar to the question that on data quality that we discussed previously.
Vidhi: So right now it does not have, but based on the additional initiatives that we are working on data quality itself in tandem to that. I'll we have plans to put it on the road map, but right now it's not there. Great. One
Chris: more question. Can Rio be [00:34:00] restricted based on access? So like for metadata security basis.
Suchen: Yeah, I can answer that. Absolutely. That's one of the reason that, we use is the underlying API. So if I have access to only say organization data, and I asked question like, Hey, show me all the individual, which I do not have access to, you will get exactly the same result that you would get if you try to go and search for individual.
Suchen: Outside of RIA, which is zero result, right? Even though there is individual records in that. So all the metadata security, all the permissions, everything that you you have applied for your tenant, for your users, they all will be honored by RENTI by RIA.
Chris: Great. That's all the yeah, that's great. So will RIA be integrated to search and match APIs for consistency? It's a good question.
Vidhi: Yeah, that's on the road. It's on
Chris: Roadmap. [00:35:00] Yeah. Exciting stuff. All right, keep going.
Vidhi: And Chris, like, when I'm saying a lot of things on the Roadmap, yes, definitely. This is just the beginning for RIA, like we discussed as well, that this is just the first phase of it, and there is a lot more that we are doing, and it's evolving as we speak as well.
Vidhi: The team is continuously building more additional capabilities on RIA. So it's not something that's very far into the future. In fact, we can also plan to have another session where we can discuss when the users can expect what capabilities are free are coming out. So that would be, I think, a very well exciting thing to know about ideas, capabilities.
Vidhi: It's not very far into the future. When I say about the roadmap, because that's how typically we assume most of roadmaps as well. So that's definitely on, on the cars is planned.
Chris: Awesome. I'm looking forward to it. Exciting. Super exciting.
Vidhi: Okay.
Vidhi: Chris, any any question that we might have left or is there something that needs to be addressed before I go to the last one?
Chris: Yeah, I just got a last one. There's no other questions at the moment. [00:36:00] Okay.
Vidhi: All right. So I discussed on this before a little bit before as well about RIA, the one that we showed and that the one that I discussed is on the documentation service and how we can, we are going to enhance the self help part of ARIA as well. Thank you. But the additional skills on RIA is going to be on is going to be focusing on advanced capabilities.
Vidhi: The one that we are going to see immediately in a couple of weeks from now, the very first week of April is we are ensuring to give a complete 360 degree view of data domains. And the one that we are starting with is customer data domain, which is going to be answering questions and helping the users with either the customer segment segmentation or Not helping in designing those market campaigns, activating customers, doing those analysis on churn analysis or whatnot.
Vidhi: So this is what is planned and all of this will get unlocked. [00:37:00] So any of those advanced fields that we are talking in context of RIA will be evident, will come to life with R for customer 360 data product that would be out in the first week of April. So that's something which would be very exciting thing to discuss next on R.
Vidhi: And while we do that, there are definitely going to be more discussions around that topic as we approach the timeline. I wanted to really make sure to come here and discuss with all of you about. RIA's is the first time we are discussing on RIA as well. So I wanted to have that discussion today to set the base for what's coming up next.
Vidhi: And I look forward to discussing it in detail in the next webinar with all of you. So with this, I've come to the end. Great.
Chris: Are there any other questions for Vidhi and team? We're really excited to rule this out and certainly the capabilities that are to come in the not so far future. So April, May we'll definitely have Fiddy back on probably in late April and or early May is my thinking.
Chris: And any other questions before we let her [00:38:00] go? Can RIA be used for invoking post APIs, not just a get APIs? No problem. Yeah.
Suchen: Oh, I can. I can answer that. So right now it is basically you're just searching for information, right? So there is still the form to fill, which RIA can basically take back.
Suchen: But that is definitely something that what we are thinking there is you can basically say I'm looking to add or create a new organization with name XYZ at 101 Main Street. So it should be able to basically, auto fill some of those information based on the mapping.
Suchen: It will understand that the name ABC seems like an organization name. 101 Main Street is an address, so you don't have to basically open a new organization, try to find that attribute that you want to basically fill up. Those kind of thing it will be able to do. But it won't post data on your behalf, because the confirmation has to come from the user.
Suchen: So at this point of time, it is [00:39:00] more of a searching. So get is what we support today, but there's a lot of exciting stuff that we are planning here which sort of relates to the question that you're asking here and it can do a lot of more complex thing in the future. Which is, one of the example that I talked about, which is finding the attribute or finding the right attribute to the right place that can be very challenging for new users.
Suchen: So it will support those things, but not right now.
Chris: So there's a question about licensing and is it included in the subscription with additional costs? Who wants to answer that?
Suchen: Yeah, I think I I answered that question sometime back, which was, uh, at this point of time, RIA document search or the article search and in the future we'll have this, uh, ability to search the community articles, post, blog post, and all those kind of thing that is included in the base subscription with your MDM subscription and the rest of the thing that we demonstrated, which is creating chart on the fly [00:40:00] and adding that to the dashboard so that other users can take advantage of that.
Suchen: Finding records using complex queries and those kinds of things that is still being developed. It is still being evolved. We thought we'll show you the sort of sneak peek of what it will look like. So once we complete that evaluation, once we fully develop it, that's when we would know more details about the cost and all.
Vidhi: Cool. Yeah. All right. Something we can discuss as we reach that path. So we, I know that this needs we need to address this question. And once we have more clarity on this we'll get back on this question as well.
Chris: Yep. All right. Suchin, Vidhi and Reltio team, thank you so much for providing us some sneak peek behind the scenes type stuff that's coming in the not so far future.
Chris: Or so, and thank you everyone for coming. Hopefully this is exciting. We do have a show coming up next week about the UI that everybody's interested in. So invite your data stewards and others that that are using the product every day and all day. So [00:41:00] until next week.