Reltio Connect

 View Only

Ask Me Anything Automated Unmerge - Show

By Chris Detzel posted 03-20-2024 09:00


Download the PPT: Automated Unmerge PPT

Last time we had a deep drive on Automatic Unmerging: Suchen showed this feature and how works, its benefits, and how it can be applied to your data management practices. There were several questions not answered, so in this show we will answer those questions live and answer ones that come up during the live session. Interactive Q&A: Have pressing questions or unique challenges?

This is your opportunity to ask anything and get expert insights tailored to your needs Connect with Peers: Engage with other community members facing similar challenges and share experiences and solutions.

Transcript here: 

Chris: I'm going to stop share my screen real quick. I'm going to let talk a little bit about. The automated unmerge and what we talked about a couple of weeks ago. [00:02:00] 

Suchen: Sorry, guys, can you guys hear me? Yeah. All right. Okay. And I'm sharing my screen here. I thought I'll do a quick recap of what this feature is about what we talked about in the previous webinar and.

Suchen: Then we can go to the question that was asked. I try to answer some of them and I'll just read through them and, go through them so that we can have a discussion or those questions and. Take any new questions that you might have. The automated unmerge, this is something that we released a couple of releases back, and we are going, more controlled fashion here for automated unmerge, what this functionality does is basically automatically detects the entities that are eligible for unmerge.

Suchen: Because the underlying merge record do not meet the matching requirement anymore, and they should be like, automatically unmerged. So it detects such records. It emerges [00:03:00] those records that do not meet the matching criteria anymore. So that's what this does. Of course the benefit of this one is, a lot of manual work can be eliminated which you are required to do today by locating those records and then un merging and things like that.

Suchen: Then we talked about some of the high level how it works and what the key benefits, things like that. We just skip that. I showed up a short demo on this 1 and I do have the same record still in the tenant, which we can go over if required. And we also talked about. So the technical configuration and usage kind of thing.

Suchen: So this is something I want to reiterate. What, like I mentioned that we started in 2023. 3 with a very basic functionality, which which is just to detect anything that is eligible for unmerge and just unmerge it, right? Something like even the manually merge record, if they do not align with automatic match rule, [00:04:00] they will get unmerged as well.

Suchen: That was in 2023. 3 And we, in 24. 1, we introduced more controls that allows you to introduce a retained conditions, which allowed you to basically tell system that, you are okay to unmerge this, but do not unmerge that. So that was something that we introduced in 24.

Suchen: 1. One of those things that we introduced in the retained condition was manually merged records. If you do not want the manually merged records. To be unmerged, then you can, or basically right now it is a default configuration. So we do not unmerge, manually merge record by default. You can override that functionality.

Suchen: You can say that you want to unmerge those records by configuring it that way. Then we also enhance the batch unmerge entities. What this job does is basically it scans through your tenant and it again detects and unmerges the record based on your configuration. Now the real time unmerge [00:05:00] works only on the data change.

Suchen: If there is a record change, then the unmerge will be evaluated only for that data change. So this batch job allows you to scan through the entire tenant and see if there are any records that needs to be unmerged. And we also introduced the reporting mode here, which just does the Reporting of the count and not go through the unmerge activity or action.

Suchen: So these are some of the improvements that we did in 24. 1. So this is how the configuration looks like. You can see here, there is a unmerge section. That is all you have to put. And enable equals to true. And it basically. We'll start detecting an unmerging. You have this retained conditions that you can put in for retaining based on the source system, retaining based on the manually merged or not, and retaining based on any additional match rule that you want us to consider.

Suchen: Yeah, this was just a demonstration of 2 records basically coming together or 3 here in this [00:06:00] case, and then 1 of the record changes name or something like that. I think that's what, yeah, here you can see Bob and Robby and one of the record was changed to David and yeah, sorry, Mark two. And that was a much automatically.

Suchen: So that was a demonstration that we had. So I will go up to the question section here.

Suchen: Okay. So these are the questions that we could not answer in the last call. So we thought we'll have this session and provide this option for us to discuss and answer some of these questions. What happens in case a business name actually change and the name changes real change.

Suchen: Ideally, 1 would want to keep the merge record as it is. I think this is a choice here, right? If you do not want such unmerges to happen, then you can introduce. A rule which will basically Say that if record [00:07:00] matches on something else, but differs on the name you still want us to keep that otherwise Unmerge will see this as a Organization name change and if organization name is must have for automatic merges, then it will be automatically unmerged so the short answer is you can control it, to keep the business name even if it is different from You The merged record, right?

Suchen: And I can demonstrate how it can be done. The next 1, is there a corresponding potential and much similar to potential match? We currently don't have that. And we briefly talked about this in the last call where I said what we are thinking about is basically introducing a new a merge. Kind of rule where you can define the condition and you can say, do not automatically unmerge this one.

Suchen: Just flag it for me or create a task for me or do something which will allow me to search for such records and decide whether you [00:08:00] want to go through the merge or not accept the merge or reject the unmerge. Currently, we don't have this functionality, but we do have a plan to introduce this one in near future.

Suchen: Can you prevent the auto unmerge applied to manual merges? So by default manually merged records do not do not get unmerged, right? And you can control this one as well. Are you considering adding discrete reports on the UI? This is similar to what I just answered. I'm going to go keep going down.

Suchen: How would this unmerge work for different values in a different crosswalk? As you have two entities where they merge based on the OV true, but we change some contributor data that isn't OV true. Would that still trigger an unmerge? I want to explain this how the unmerge works, right? Let's go to this record.

Suchen: Let's look at the crosswalk. So what [00:09:00] it does basically, it doesn't matter whether the data was changed or the OV value was changed or not. Anytime there is an update to the record, it looks at all these contributors. In this case, there are two. This is not seen as a contributor because this is just a cleanse function data, which was generated by Reltio.

Suchen: So this one is seen as a crosswalk or contributor in this one is seen as another contributor. So we just take this to contributor. Treat them as independent entities in memory and basically we just run the matching on this one and they have their respective OV values, right? When they are pulled apart.

Suchen: When we run them through the matching and if the result come back as they should be merged using one of the automatic match rule, then they stay together. If after going through all the match rules. If it does not match on any of the automatic match rules, then they get pulled [00:10:00] apart. That's how it works.

Suchen: So going back to this question it will depend on, what the OV value is going to be. When you look at that one individual contributor. 

Chris: Hey, quick question wouldn't the reporting effectively be the potential unmerge? Oh, 

Suchen: what's the question? Sorry. I didn't 

Chris: know. Gino mentions, wouldn't the reporting effectively be the potential unmerge?

Chris: Yes, you had a question before, 

Suchen: Hey, do we have potential unmerge? And. I was thinking that would be that is correct. Yes, but with the current functionality we have in the reporting mode is that it just provide you the count. It does not give you the complete list of URIs.

Suchen: That's why we are going to extend that, which will provide you the complete list of URIs. So that will be your phase one of potential unmerged. You can look at the list. And you can then decide whether you want to [00:11:00] unmerge or not. But the second phase will be like, a way to flag this record, for example, and just call out exactly which crosswalk is eligible to be pulled out.

Chris: Another question in the chat. If we're having two records, say A and B, both are having suspect match. We have unmerged those records and later due to some value change, it is If it is again, satisfying auto match rule, then will it match again? 

Suchen: Yes. So if you have unmerged those records, you have to make sure that I'll show you here, if you have unmerged them, they get registered here, meaning if I pull out one of this crosswalks from the UI or using API, using this unmerged functionality.

Suchen: This will become a separate entity with this entity id. And you will see that under this not a [00:12:00] match list here, it'll be listed here. What it does by default when you do it from the UI or by data steward, it keeps those record apart forever. The idea there was data steward has reviewed this one.

Suchen: And decided that this to record another same person or same organization or whatever that entity type is, and they should be different and it stays here forever. But if you change your mind, there is a action item or action button here, which allows you to reset the matching. You can say. Reset this 1 and see if it matches with any other entity or with sorry, with the same entity again, because something has changed.

Suchen: Unless you reset that, it won't automatically merge again. But if you do reset that, then it basically becomes eligible for matching again.

Suchen: Yeah, thank 

Chris: you. So another question. When two records are merged [00:13:00] after but after some update these are automatically unmerged Will the update time change for both of the records? 

Suchen: Yes, because So automated unmerge, does not have any separate functionality This is just basically detection and then triggering the already existing unmerged functionality So today's unmerged functionality when you something you are technically changing the record, right?

Suchen: Right now, if I pull this crosswalk out. This crosswalk could be taking away the OV value of this consolidated record. So the last update date on both these records will change.

Suchen: Thank you. I think that's it for that. Okay. Alright. Let me keep going here. How about those entities? Okay. So this was basically talking about profile with two [00:14:00] addresses and profile two with one address, which is same as profile. One of the profile after a split or unmerged the common address needs to be split to who will get the original ID who will get a new ID.

Suchen: Again, going back to this record here, let's say this record is Mike car wash and the other my car wash had multiple addresses when we unmerge. We maintain the link between their original association with the addresses. So if the first address came from my car wash and the second and the third came from Guitar Center, then those addresses will be unmerged as well.

Suchen: And they will go to their respective original entities. But if you have created a new address after they were unmerged, after they were merged. That new address always gets attached to this entity ID, which is a surviving ID So I can see that it's this one. So it will go with the my [00:15:00] car wash So that is only the new record that gets added On top of this one, right?

Suchen: But if you explicitly tie to this crosswalk Then oh, sorry this one then even after unmerge it will go With the guitar center here. Okay, great. 

Chris: We have a couple more questions in the chat. So do we have some sort of a report that shows all the records that were unmerged as part of the recent data load?

Suchen: So you have this bulk and merge in the report mode that you can run. I love to check whether we have a date that we can provide to run the report like, for only those record that was changed. Between certain time range or whatnot? Yeah. Not then that is something that we, yeah.

Suchen: I think you can just query that through the audit log. You search for un merges by the person who loaded the file within a time range, [00:16:00] right? No. I think the question was if I load some data, which changes, out of 15 million records, it changes only 10,000 records. Yeah. How do I get the report on those only 10, 000 or something that was changed in.

Suchen: Just last 24 hours or whatever, if I want to run the report only on that. So my bulk and merge. Let's try that.

Suchen: It's an idea. Oh, yeah, I will get back to on that. But yeah, what, what will happen is The bulk and merge job that we have we can [00:17:00] run that in the timestamp and you can say Run this report only on those entities that was changed in last 24 hours or in last one month and things like that So that's how you would do it.

Chris: We have some more questions So i'm going to start asking the ones in the chat because all these have already been answered. Will it impact Reltio load date also and can you please help me to understand how Reltio load date 

Suchen: change? Yeah so we have three dates here, actually four for every crosswalk three yeah, four dates, create date, update date, arrange your load date, and delete date, right?

Suchen: If you delete this crosswalk, that's the fourth date that you will have. So we have create date, which is basically the creation when this record was created, or this crosswalk was created. Load date and update date, this is basically when the request or source system basically initiated the request and for some slowness in the platform here or the, [00:18:00] there was a large data load and this record was just sitting there in the queue for some time, which is why it got delayed here.

Suchen: Yeah, that's what happened probably here, which is what it tracks. Relative load date is always that date when the changes were applied to the primary storage and update date is basically that the date that you can provide and you can say this is when it was changed in my source system and you want to track it.

Suchen: It doesn't matter when it caught to the or LTO tenant. So those are the two dates that we allow you to provide. This is a LTO generated date, which is a system time. This is also a system time and update date is, uh, user provided or system provided date.

Chris: This one's related to unmerge or auto unmerge. What is the attribute field that we would reference to ensure. Exports get triggered correctly and downstream apps can have appropriate info propagated. [00:19:00] Okay, 

Suchen: so this is so auto unmerge like I said this is not a new event or new functionality that we have introduced on the action part.

Suchen: This is just a detection that we have introduced and we are still using the underlying unmerge functionality. Which was always in the platform. So the event that you should be looking for is the entity and merge or split and event. That's what you should be looking for to to detect that unmerge was performed.

Suchen: Okay, 

Chris: Will the automatic unmerge process mark records as not a match? 

Suchen: No. This is where we have this is the only difference in the behavior when automated unmerge does the unmerge. We reset the match so that it can immediately go and match with the right candidate if there [00:20:00] is one. If data steward does it, then it is always marked as not a match and it stays there like I mentioned until it is reset, but automated on merge will always automatically remove this not a match and it immediately becomes eligible for matching with the same record or with any other record.

Chris: Could you comment on Reltio Salesforce connector impact? So we would have deleted a merge in Salesforce. How would unmerge have affect such a entity? 

Suchen: Yeah, so this is one use case that we have not. Fully tested and we have called this out on our documentation that connected part is something that we are still working on, but that there was a customer basically wanted to try that.

Suchen: And what we have done is we have introduced this retained condition. So that if you do not want that data to be touched and you want to use this on much for everything [00:21:00] else you can provide that. So if you say salesforce. In the retained condition this will allow you to basically maintain the backward compatibility of all salesforce interaction with radio at the same time.

Suchen: Take care of the emerge of any other source system. And we'll test out the salesforce connector and other connectors that we support and provide that clarification. But at this point of time the way it would work is if you delete something then that will trigger the unmerge evaluation.

Suchen: And if it is eligible for unmerge, it will go and unmerge as well. It doesn't matter which source system it comes from unless it is not listed under the retained condition. 

Chris: So when investigating the auto unmerge results, how do you tell? Which crosswalk, the address came from? 

Suchen: So yeah basically what we are doing here.

Suchen: That's a very good question. And I [00:22:00] say that because when we were designing this 1, that was 1 of the challenges that we had. The question is, if this 1 had 3 addresses, how do I know which addresses belong to which crosswalk here? Like I said, when we merge. We always maintain the lineage between the entity and the their related entities, right?

Suchen: Which is address in this case, how they came in so that we can revert them back when the unmerge happens. So there is something called as a loser relationship id that allows us to basically grab the entity addresses that belongs to that entity. So when we uner them actually, or when we pick them for evaluation, we know exactly which addresses came from this crosswalk versus this other one.

Suchen: So yeah, we do maintain that information, which allowed us to run this evaluation on their original state.

Chris: Great. I feel like you [00:23:00] know your stuff today. That's really good. Can automatic unmerge be limited to a specific source or sources? 

Suchen: Yeah. So that was another thing. While designing, we thought we will make this as a sort of inclusion design, which means just include the source system that you want us to evaluate, right?

Suchen: But that could be a long list of source system for some many customers. If you have 13, 14 source systems, then you have to list all of them. So the way it works right now, it's exclusion, meaning list what you do not want us to evaluate and it will leave those records out. So the way you would achieve this, what you're asking here is limiting to a specific sources is by listing everything that you do not want Unmerge to touch and leave everything that you want us to consider.

Chris: How is this functionality, how does it work if match IQ is used, the same way as regular match rules. So just confirm his [00:24:00] understanding. 

Suchen: Yeah. So that's where this second retain condition comes into play. What this what this does is basically by default, when the unmerge is evaluated, it It looks at that two crosswalks and it runs through all the automatic match rule to see if, does it match on automatic rule number one.

Suchen: If the answer is no, it moves on to the second rule and check if it matches on automatic rule number two, if it doesn't match on the second one as well, and there is no other automatic match rule, it just goes and unmerge. But if you have a retained condition, which can be anything, it does not have to be automatic match rule, it can be suspect as well.

Suchen: It will evaluate this one as well in this case, because it's in a retained condition and it will evaluate this rule. And if this one comes out as true, then it does not unmerge. So what you'll do here is you will list your match IQ as one of the [00:25:00] rule here, and that will prevent it from unmerging the match IQ merge records.

Chris: Is there a way is there a way to find out how many records got unmerged? 

Suchen: Yeah, the way you would do that is let me see if I have configured that if not, then i'll just walk you guys through it. So through activity yeah, I don't have it here Maybe let's see. Yeah here you can see here.

Suchen: This one has a You It captures, when it was merged, when it was unmerged. So we track this at the end at the tenant level as well. You can go to the activity log and you can say, show me all the profiles. That was unmerged in last 24 hours or 30 days or whatever, and it will show you that list here, but there is no option to download that entity IDs from here.

Suchen: But, yeah, [00:26:00] this is how you would do it. You'll go to the activity log which you can configure on the dashboard and you can extract that list from there.

Chris: I realized we're discussing Unmerge and related to propagating data downstream. If we merge, will the same attribute fields referenced for Unmerge be used to ensure we trigger exports for downstream? 

Suchen: Let me understand this question. So basically you Are you talking about and you can feel free to unmute yourself and elaborate on this question, but you have some process built, which tracks the attribute that is used in emerging.

Suchen: And when those things change, you have the integration process. Oh, there is that. Is that right? 

Chris: 2 points, you referenced 

Suchen: earlier when we did the unmerge. In order to 

Chris: successfully inform consuming applications of a 

Suchen: [00:27:00] split that we've done, you referenced a couple of attributes, a couple of fields that we keep in on to ensure that an export occurs.

Suchen: So the downstream applications receive the correct information. So now if we flip it around and we merge records, I want it to determine. 

Chris: What is 

Suchen: attribute field that we reference to 

Chris: ensure that the downstream applications, which. Prior to 

Suchen: us merging had two separate records and now we're trying to inform the downstream applications We are now merging 

Chris: two records 

Suchen: Yeah, So there is a merge event as well.

Suchen: I think it is Let's

Suchen: look at this[00:28:00] 

Suchen: that's giving me a lot of other options here, but let me talk through this So you have this entities merged, event that will be triggered You Which will have the list of those two entities that came together as part of automatic merge, or if somebody manually merged it, so that will be your basically indicator that two records that that downstream had as independent records is now merged into one single one.

Suchen: Thank you. 

Chris: So if you exclude, retain source one, and if source two is merged with source one, then we'll unmerged. Okay. Yes, 

Suchen: So we don't yeah we evaluate okay, let me see. Source one was in retained.[00:29:00] 

Suchen: No not in this case. So basically we check to see if a record. Has at least one of the source system that is from the retained condition. We just leave that record alone unless if there was a third record source, which was from source three, let's say, and source two and source three came together and source three record was changed.

Suchen: Then that record comes out. Basically, it's like a tree, right? So you have a matched with B.

Suchen: And so this is from source one. This is from source two, and this is from source two. Again, if C changes, then we take out C. But if A changes, then we don't take out A. Basically it's trying to, [00:30:00] and by the way, it will also evaluate to see if A and B are still going to be a match of each other. If we take out C, if they're not, then we don't touch this record as well.

Suchen: In your scenario, if this was a requirement, sorry, if this is a scenario, you have source 1 and source 2, A and B record and source 1 is in retained condition, we leave this record alone. We don't touch this.

Chris: We might come back to yours, Max. I know we answered it a little bit ago, but we could still come back. How does auto unmerge evaluate manual merges? 

Suchen: Yeah, we go back here.

Suchen: Okay.

Suchen: So as you can see here, this one basically this one has it says there's a IRS admin or automatic merge tracking. So it tracks who did the merge, is it system or is it a user? [00:31:00] So if it was done by a user, if this tree has any crosswalk that was brought in by as a result of manually merged record, then we do not touch that record.

Suchen: So to answer your question, we have the history of the merges. That we look at and see if there is at least one manually much record in there, then we don't touch that record.

Chris: So going back, so I know you answered this question, but if you could answer it again, when investigating the auto on merge results, how do you tell which crosswalk the address came from again? Yeah, 

Suchen: I think I answered that and we have the lineage of. The original association of the address to their respective contributor, which is basically this ID.

Suchen: So if the, there are multiple addresses that it will be tied to either this ID or the other ID, which is a loser ID. In this [00:32:00] case is this one, because I know this is a winner ID, which matches with this ID here. So the addresses are always tied to, or any relationship attribute for that matter is always tied to their underlying.

Suchen: Original IDs. I still don't 

Chris: see how 

Suchen: you, you're doing that. How are you seeing that it's associated with the original ID? Let me explain it here. E1 is my entity ID and I have my address 1. I have my address 2 again, and this is my E1. And let's say this is my address 3. So this E1 represents the URI of the entity 1 and E2 represents the URI of entity 2 and A1 represents the URI of the location ID.

Suchen: And this is again, A2 and A3. So what we have is when you have E1 [00:33:00] associated with A1, we have R1 here, which is a relationship ID. And then we similarly have, we have R2 and this one will be R3 because relationship IDs are always unique across. So when this record get merged, the data that you will see in the system will look something like this.

Suchen: Now, this will change to E1, let's say, because E1 survives. But you will notice that this one does not change here, right? Okay. In this case, this will be one second. This will be R4

Suchen: and there is R3 with A3. And this is E2, and this will be in the deactivated kind of,[00:34:00] 

Suchen: let me like that. So this will be deactivated, but we still keep track of it. So when we look at the unmerge, we look at this deactivated record as well, or this association. That's how we know that. A3 is part of E2, which is what we have here, which is this ID here. So that's how we know and you can get this information too.

Suchen: If you do get relationship on this loser ID, it will tell you that it is originally associated with E2. Can you see those relationship IDs in the interface anywhere? Not by interface. Yeah, not the loser one. You will only see the winner one. And how do you do that? Yeah Let me activate that

Suchen: It's in the wrong tenant. [00:35:00] Okay, there it is. Let's go to organization here and I'm going to activate this wait, give it a minute there and i'll come back to it So you will see the relationship ids. As one of the tabs here and that's what you will see but you would not see the deleted ids or The inactivated IDs, and those are soft deleted one.

Suchen: They're not like hard deleted. So we can still access them. Okay. All 

Chris: That's helpful. Thank you.

Chris: I wanted to check if we are building new, if we were building a new project set up on relative, then do we have any checklist or documentation, which we can refer to so that we're not going to miss anything on requirements, gathering merges, unmerging survivorship. Configuring, relative set up.

Suchen: Is there any, yeah. Yeah, so we do have some guidelines, some checklists and what I would suggest is you can just ask this question on the community one of our peers, [00:36:00] our LTO, our professional services team members should be able to point you to that or reach out to you and help you with that.

Chris: Yeah, I have something that I'm looking for now that I think is going to be helpful. Sarah did a show a couple of weeks ago, 

Suchen: yeah, okay, I just want to go back to that previous question and I think the deployment or the publish is done. So we should see the relationships here. And so this is the address and you will see the reference, sorry, the relationship ID is over here.

Suchen: That's your relationship ID.

Suchen: Yeah as you can see since it is coming from this both source system. Or basically those two different IDs, it is maintaining those different relationship ID. That's how we know both those records have the same or pointing to the same entity. A location entity, which is this address here.

Suchen: [00:37:00] So this is that idea that I was talking about, which is R1 R2.

Suchen: Thank you.

Suchen: All right. I would to go back here and just talk about 1 of the thing that we have. That we are working on so retain manual merges is like I said, it is true by default, which means you, we will always retain the manually merge records. We will not touch that. We will not evaluate unless this property is set to false here.

Suchen: And if you want to undo everything that data stewards have done for whatever reason, and you want to go back to how it was before a data stewards started manually merging them, you can set this to true, sorry, false. And only [00:38:00] those records which meets the automatic match condition will stay together and everything else will be pulled apart.

Suchen: So we are working on similar functionality for merge on the fly or merge by crosswalk functionality. What that means is you can merge records without using match rule by simply providing the crosswalk of already existing record. Which means if I were to, if I had just this record. And let's say this record was nothing or is not similar to this other record, but For some reason, you know that this is the same record as the other one, you can completely bypass the matching engine and simply provide the source system information of the existing record in the request body of this record here.

Suchen: The contributor equals to set to true for this using this record. So basically it's a lookup kind of thing. Look [00:39:00] up this record using this crosswalk information and simply attach this new crosswalk to that. That's what it means which is merging by crosswalk. Today, those records are unmerged by default.

Suchen: And we realize that is also should be treated as a manual merge. It is just that humans not doing it, it was some system or some integration layers doing it, right? So we are going to introduce that and that is also something that will be available very shortly. Is it 

Chris: possible to search in the Reltio UI  to I with relationship.

Chris: I. D. 

Suchen: no, you cannot search on the relationship ID. But you can search if you configure that reference attribute not. Not the ID though, but like this one is a reference attribute, sorry related entity. So I've configured that as a reference attribute. Now I can use the attribute of this related entity and [00:40:00] search it using this reference attribute, but you cannot use the ID itself, which is our relationship ID for searching on the UI.

Chris: Okay so can we configure this manual merge retention using Reltio UI? 

Suchen: No. That is also something that we currently do not support. This can be done only through L3 at this point of time. That is something we will also support through this data modeler in the near future. So you will see somewhere go to the data entity types.

Suchen: We are currently working with organization sorry. And you will basically go and click here. This is where you will see that option to unmerge a configuration, right? So it's not in this Application yet, but that is something that we'll be adding in the near future. 

Chris: Okay. Do you have a release date to unmerge by crosswalk that you just described?

Suchen: Yes, that is, uh, before 20 [00:41:00] 24, 2 June. In fact, actually, we might be able to release it much sooner than that in a couple of weeks or maybe even sooner than that. It's they'll get us 

Chris: so excited, man. . Yeah, . 

Suchen: Okay. Yeah, we're, like I said this is we started in 23.3 with a very basic version.

Suchen: We added more controls. We are going to add more controls. The idea here is to give you give our customers the flexibility to choose the data set that you want to use on much for, so you'll see more feature coming in into this into this feature. Yeah, 

Chris: there's a lot of interest in this.

Chris: And, I'd be very interested in the next few weeks, month, two, how many of folks that are on the call today. Actually using auto unmerge and then going deeper into that. So you, maybe you and I can, maybe it's May or June go a little bit deeper into this and what's come 

Suchen: out on that.

Suchen: Yeah. And using this combination of retained condition, this already gives you a lot of flexibility. So imagine if [00:42:00] you want to unmerge just one source system at this point of time, you can list. All of the source systems here and then you can also list all the additional match rules that you do not want us to consider and you can go after a very tiny list of entities just for testing purpose and then slowly take out this one source system at a time and expand that, you know, and move on.

Suchen: Move around and cover more and more use cases or let unmerged handle more and more sources systems, right? So that is the idea here and more written condition will come. One other candidate that, that is definitely looking very good to be in the near future is a potential unmerged, which you might have guessed that there's a lot of interest there.

Suchen: ability to see that from the UI for customers to review them and decide whether to accept the merge or not, a lot of stuff to be added. 

Chris: A [00:43:00] lot of great stuff. And this wasn't even a demo and you showed us a bunch of stuff on the system here. So it was really cool. Really appreciate it. Is there any other questions before we, or any other thoughts switch in before we 

Suchen: yeah, we'll just open it up and see if anybody wants to ask anything. Otherwise yeah, I'll just summarize what is coming next or what is that you've seen other customers do? I can just spend five minutes, talk about that. Yeah. Let's do that.

Suchen: Yeah. Feel free to ask any questions that you might have. If you don't want to type in the chat, you can feel free to ask. And yeah, there's no more 

Chris: questions at the moment, but 

Suchen: okay. Like I said, of the customers are doing right now is just list all the source systems that you're currently loading data for and take out any of the 1 of the source system, which is really small data set or whatnot and try to build a a [00:44:00] rule.

Suchen: In addition to the automatic match rule and this could be a potential match like a suspect. It's not going to do or it is not going to have any impact on your merge outcome or the match outcome, right? It is just going to generate some additional potential matches. You can list them here. So this just gives you a control set that you can play around with.

Suchen: But I would definitely encourage. Do not go with enable true and leave all of this out. It's just going to go after your entire data set, right? Just to test it out, have as many written condition as you can get comfortable with the feature. Let us know if there is something that we can add, improve in the existing written condition and we can improve from there.

Suchen: And if you feel comfortable with that tiny set of data that we are unmerging it correctly, but not, then you can slowly expand. And start taking out stuff from the retain merge condition, which [00:45:00] means you're allowing automated and merge to take on more work and let data steward basically focus on something else.

Suchen: I would definitely encourage everyone to try this and let us know the feedback, if there are any issues, any improvements that we can in addition to what we have already discussed. And all these questions that was asked today, we'll try to create a request. We'll invite everyone to come in and vote on those because that will help us understand.

Suchen: What is that we should go after first based on the voting system? What is the number one priority for most of our customers? So we'll look out for that, or you can go and create your own idea to get as great. Yep. 

Chris: All right. Suchan, thank you so much. We'll be seeing you again in the next month or so once we, you and I collaborate a little bit more on the next show on something a little different.

Chris: But until then, thank you everyone for coming to today's Ask Me [00:46:00] Anything.