Find the Reltio Integration Hub Demo and Presentation here.
Transcript Below:
Chris Detzel (00:00:07):
For those that don't know me, my name is Chris Detzel and I'm a director of customer community and engagement. And we have our special guest, Ayan Basu. He's the senior director of product management, and he's going to be joining us June and July for lots of shows. So I'm really excited for him to present on Reltio Integration Hub, Building Recipes as Easy as Pie. Ayan came up with that, I think. We are excited to bring this back. I know we talked a little bit about this several months ago, but Ayan's gat a lot of really cool stuff to show today, and I'm really excited to share it with you. So the rules of the road are please keep yourself on mute. All questions should be asked in the chat or take yourself off mute, and you can ask directly. I will make sure that these questions are asked to Ayan.
Chris Detzel (00:01:04):
Community show will be recorded and posted to the Reltio community. The hope is by the end of this week. Now, we have a full month of this month and next month of community shows. I'm going to just tell you what they are real quick. Today is the Reltio Integration Hub. Again, really excited about this. Then tomorrow we have a show on how to build an MDM practice from ground up. Chris Iloff will be helping us with that. It's a little bit different. It's not product-focused or related. It's about the practice of your MDM. And then we have, it is a Reltio-focused, the new modern Reltio user experience. That's coming up. And the Reltio CCAB, that's kind of a thing that is for specific folks that are interested in helping with our documentation and things like that.
Chris Detzel (00:02:01):
And then we have a show on UI configuration self-service. And then one another show that isn't Reltio specific on the 19th is natural language processing, text representations in MDM. So I'm very excited to show that and share that. And then moving data in and out of Reltio, it's a cake walk. It's going to be really fun. Machine learning based matching. That will be a Reltio focused. We're bringing that back. We were supposed to do that several months ago and something happened, but now bringing it back and we have some cool stuff there. And then lastly, Ayan will be presenting with one of our customers, Sajid Sayed at Advarra around Reltio Integration Hub, and that's going to be really fun to do on the 28th. So lots of really good stuff for the rest of the summer to show you. So if you haven't RSVPed, just go to community.Reltio.com/events and RSVP today. And I've done enough talking. Ayan, it's all you, man.
Ayan Basu (00:03:05):
Awesome. Thank you, Chris. Hi, everybody. Ayan Basu, senior director of product. Very excited to be here and thank you for joining today and thank you for your time. So what I'm going to do is share my screen and walk you through the relative Integration Hub. And so for the agenda today, real quick, I'm dividing this up into three pieces. First is giving you that overview of Reltio Integration Hub. I have a few slides I want to walk you through. There's also a lot of customer momentum around this. So we have real life examples of how customers are using this product. And then lastly, I'll do a demo as well. You might be wondering what a recipe is. So I'll show you what that means and how it can benefit you and what you're trying to do. So lots of really great stuff coming and really looking forward to spending this time with you.
Ayan Basu (00:03:57):
So without further ado, I'm going to jump into this. So what I wanted to do is tell you a story of Joe here, who just started at his brand new company, Smith Corporation. And he's been tasked to roll out this MDM implementation at this company. And they recently purchased Reltio MDM. And one of the things that he's really thinking about that he is really concerned with is how does he build all these integrations to different systems? So if you think about Reltio MDM, it's connecting to all these different applications within their technical ecosystem. And Joe can do it. He's done it many times before. He can write these applications, run it somewhere. But it doesn't just end at one, two or three, that number of applications that he has to integrate to. It is something that grows over time. So that's where it does become a challenge to Joe.
Ayan Basu (00:04:55):
And he's trying to figure out how to do this in a seamless manner. And this is an actual like industry-wide trend. And this might be something that resonates with you where it's not just a data sprawl that's happening across enterprises. It's also an application sprawl. Every single enterprise is dealing with hundreds, if not thousands of SaaS applications and there's brand new applications being introduced day in and day out. So this is a problem that is exponentially increasing in size. And so if you think about Reltio, what we do is get you that single source of truth. We master data across all of your applications. We get that data and provide you that golden record. But then at the heart of it, it's integrations and making those integrations with all these applications seamless and easily integrated, interoperable and also automation, which is the big thing.
Ayan Basu (00:05:52):
So all of these things become super critical to an actual, full-blown data initiative. And today what's happening is the status quo way of creating these integrations, running these applications, writing hundreds, if not thousands of lines of code, it's holding organizations back. And that's because creating these one off integrations, it's cumbersome, it's unreliable. These are point to point integrations. It's difficult to manage, they're time consuming. Every integration requires an IT team. It's a development project, it's expensive. And then lastly, it creates an inherent rigidity in your organization where you're unable to be flexible and agile, especially as new applications come on board. So with all these things in mind, these are the problems that we were looking to solve. And that's where Reltio Integration Hub comes into the picture. And it's our all new low-code/no-code platform that we introduced a few months ago.
Ayan Basu (00:06:52):
And what this does is it gives you the flexibility and organizational agility to be able to connect to whatever application that you need to, and you can solve whatever use case you need to solve. And on top of that, we have thousands of out of the box connectors that you can use. So if you think about low-code/no-code, and the fact that there's so many out of the box connectors, you can imagine how that drastically speeds up your time to value and how quickly you can get these integrations up and running. So we're very excited to launch this product and that's what's bringing us here today is where I can actually demonstrate to you how this product works and the value that it provides. So in keeping with that theme of quick time to value, there's a few different considerations here. So one of the value propositions of this tool is it's cloud native, it's real-time, highly performant, highly available, everything you would expect of a modern SaaS platform.
Ayan Basu (00:07:52):
Then the second piece is the agility where we do have all of these out of the box connectors. And then the last thing is ease of use. It's drag and drop, no-code/low-code. So you can create these integrations in a simple, performant manner. And it can be done in minutes as opposed to weeks, months, years, what have you. So just double clicking on the real-time operational aspect of it, as I mentioned, highly available performance, scalable platform, world class security standards, all the data is masked as well. So unauthorized users will not have access to this data. And we've seen that you can load hundreds of thousands of records within minutes. So from a performance perspective, you don't have anything to worry about with this platform. Then on the business agility side, I talked about the thousands of connectors we have. But then on top of that, it's low-code/no-code, meaning you can drag and drop these boxes.
Ayan Basu (00:08:50):
You can create these workflows. You can create integrations, you can orchestrate various APIs. You can transform data if you need to. And all of these things you can fully automate. So these can run behind the scenes, whether it's a batch job or a real-time job, however you want to have these enabled, it's completely up to you how you want to run these. And there's an immense amount of flexibility in terms of the way you can build. Your use cases can be as simple as you need them to be, or they can be enterprise grade. Again, completely up to you in terms of how you use this to solve your challenging business problems. Then we do have this concept of prebuilt recipes. And a recipe, the best way to think about it is a workflow or an integration that you're creating using that drag and draw functionality.
Ayan Basu (00:09:40):
And what we've done is we've, pre-built a bunch of recipes, which I'll show you when I show you the product. And with these recipes, you can actually just use them and customize them for your own use case. So you don't have to create something from scratch. You could actually take something that we've built and just make a few tweaks to it and get it running for your own specific use case. And this is where you can be up and running within minutes, because you're already starting off from a template as opposed to starting from a blank slate. So that gives you a lot of power and a lot of accelerated development ability here. I did want to take some time and show you how our customers are using this product today. So we have a lot of customers that are looking at it, a lot of interesting use cases.
Ayan Basu (00:10:30):
So I'm going to jump right into the first one. And this is a customer that is actually a reference customer for us. And we just put out our press release today, a couple of hours ago, in fact, where Advarra was highlighted and they had a quote in the press release where Reltio Integration Hub actually tremendously helped them accelerate their development. So overall, using Reltio Integration Hub, they were able to do it three times faster to integrate. And also, they saved around a $100,000 as a result of it. So the business problem here is they needed a way to quickly stand up their MDM. And a key part of that was getting data from Snowflake into Reltio. And they had to do this in a short amount of time. So as you can imagine, without Reltio Integration Hub, the only option that Advarra had was to have custom code written, hosted somewhere externally, and then have that connect Snowflake to Reltio.
Ayan Basu (00:11:32):
So having to build something fairly complex, to be able to do that, and that's what would have costed them about $100,000 dollars. So instead of doing that, they use Reltio Integration Hub. They built this integration, this workflow within about six weeks. They saved a lot of consulting hours. They saved having to stand up a whole team, having to stand up additional infrastructure. So all of this was self-contained within Reltio using Reltio Integration Hub. So that's where the value really was readily available to them. And they saw how something like Reltio Integration Hub solved their real problem that they were facing. So that's one. Then another one that I'm going to talk about here is, one of our customers needed to get customer accounts into their system and understand the relationship of their customers, whether they were a dealer or a distributor, and understand the nuances there.
Ayan Basu (00:12:31):
So they used event queues to be able to do that. And they loaded about 150,000 records and created 250,000 profiles all under a couple of hours. And there were some fairly complex logic in there. There were transformations in there. And using all of that logic, using on Reltio Integration Hub, they still experienced really good performance using event queues. And the quotes that you see up here, they actually said these on call, where they found the tool to be ridiculously impressive. They found it to be much easier and simpler. And in fact, the partner that the customer was working with, they mentioned that if Reltio Integration Hub was not in the picture, they actually would have to have written about 5,000 lines of code to do the exact same thing that Reltio Integration Hub was able to help them do, just using a drag and drop process.
Ayan Basu (00:13:29):
So then with another customer, this is where they wanted to integrate to a data stewardship app called Matchbook. And what that does is it gets data from Dun & Bradstreet, and then there's a data stewardship capability on this tool. And they needed to integrate that stewarded data from Dun & Bradstreet, back into Reltio and enrich Reltio records. So this customer was able to save about 100 hours of manual processing a month that was happening as a result of integrating the data stewardship app with Reltio, using Reltio Integration Hub. And so once everything was said and done, this is actually running behind the scenes. 10,000 records are being processed on a weekly basis. And these jobs are automated, which means this gets rid of the manual aspect of having to do this. So this is all automated, they're improving their productivity, they're improving their operational efficiency. So this is real value being provided by the Reltio Integration Hub.
Ayan Basu (00:14:35):
In this example, we had a customer that needed to enrich their data within Reltio using US Census information. So US Census data is available publicly on the US Census Bureau website. So using the Integration Hub, we were able to create a connector to those APIs that are freely available and enrich the data in Reltio with the Census data that this customer was needing. And this is something that they were struggling with, they weren't exactly sure how they were going to do this. But thanks to the Integration Hub, we were able to work with them and demonstrate viability of doing this integration and enriching their data in Reltio within about a week. And right now we're currently going through the process of getting this into production. So again, another really exciting use case that the customer is able to solve.
Ayan Basu (00:15:30):
Then in this last big example here, I wanted to talk about what this customer was doing using Kafka topics. So the way their system works is they have all of these marketing leads that are in, whether it's SFDC or homegrown applications or marketing campaigns. They're getting all of that lead data, and it's all going into a Kafka topic. And so with a out of the box confluent connector that's available on Reltio Integration Hub, they were able to get all these messages that are constantly being streamed in, and they built the recipe using this confluent adapter integration and getting all of that data through into Reltio. But in between the Kafka topic and Reltio, they also put in some pretty complex logic and data quality rules where they're checking the leads to make sure that the first name is good, the last name and email and address, all of that is meeting their data quality standards.
Ayan Basu (00:16:32):
So by building all this logic, they were able to check all of these messages that are coming in, check the values, and then only sending the accurate, good information into Reltio. And they were able to do this within about a four week timeframe. And now that it's in production, they're processing about 10,000 leads on a daily basis. And they're getting Reltio completely synchronized with all of their systems within their organization. So pretty interesting. And this one, the customer actually was able to do it by themselves, which speaks to the self-serve capability of this. So there wasn't this whole long, drawn out enablement process. The customer looked at it, it made intuitive sense. They looked at documentation and they could actually figure this out for the most part themselves. So that's another big value driver of this tool is the self-service aspect of it, and the fact that you can get up and running without having to go through months of training courses, and really trying to understand how this works.
Ayan Basu (00:17:36):
So simplicity of use is a key value proposition of this tool. So to end to the customer piece here, I wanted to quickly cover off on a few other use cases that our customers are currently leveraging the Integration Hub for. So in the retail industry, we have a restaurant chain that's looking to replicate menu items across multiple regions, and they're using Integration Hub to do that. Within the hospitality space, a couple of our customers are currently working on integrating to SAP SP4 HANA, as well as additional use cases around Snowflake integration and then on-prem Excel. In the life sciences space, we have customers integrating to Veeva CRM as well as Treasure Data, CVent. And then also another customer that's looking to really speed up data loading from a bunch of different sources and simplifying and accelerating that using the Integration Hub.
Ayan Basu (00:18:37):
And in the finance industry, again, a couple of use cases around data loading from JSON files. And then lastly, some use cases around integrating to NetSuite, as well as integrating the BvD connector. The BvD connector is something that we offer out of the box. It's something that we built on the Reltio Integration Hub. And it's a data enrichment connector that exists just like Dun & Bradstreet or another data provider. So as you can see, very flexible, expansive use cases. Customers are really using this across the board to solve whatever problem they're running into. So a lot of excitement around this, a lot of interest from customers. And then before I jump into the actual product demo, I did want to pause and see if there's any questions coming through, anything that I can answer right now.
Chris Detzel (00:19:30):
As of now, I don't see any questions in the chat. But if somebody has a question, feel free to speak up and/or put it into the chat.
Ayan Basu (00:19:41):
All right. Feel free to ask questions then, chat. And then Chris will read them out and I can answer them.
Chris Detzel (00:19:48):
There is one question, sorry. Reltio Integration Hub can be used to publish the data. Is that true?
Ayan Basu (00:19:58):
Yes. The way Reltio Integration Hub can work is, you can either get data out of Reltio into another system. So Snowflake is an example. You get data out of Reltio, put it into Snowflake or Redshift or Google BigQuery, wherever you want it. Or you can get data from a Snowflake or a data warehouse into Reltio. So it works both ways.
Chris Detzel (00:20:24):
Thank you.
Ayan Basu (00:20:25):
Sure. Feel free to ask questions. Definitely happy to answer them. And while we-
Chris Detzel (00:20:33):
We do have a couple on starting to come in now. I knew this going to happen.
Ayan Basu (00:20:38):
No worries.
Chris Detzel (00:20:39):
Can I use Reltio Integration Hub instead of ETL?
Ayan Basu (00:20:45):
It really depends the use case. Reltio Integration Hub, it's designed to be a flexible tool with a lot of utility. It's not necessarily purpose built to be an ETL. Although we do have customers that are using it in a similar capacity. What Reltio Integration Hub does provide is the ability to transform your data in a low-code/no-code manner. So given the options that are in the tool, the transformation capabilities, you could potentially do a ETL process within the Integration Hub. However, as I mentioned, it's not purpose-built for that. So it really comes down to your use case and what you're trying to accomplish and comparing the capabilities of the Integration Hub to, say, another ETL tool that you might be considering.
Chris Detzel (00:21:36):
And this question comes up a lot, so I'm sure you're ready for it. Why use the Salesforce connector or Snowflake connector when Reltio Integration Hub is available? Any pros, cons for both methods?
Ayan Basu (00:21:48):
Great question. That question does come up and we will be looking to address this in a future webinar. But the answer right now is, we have these bespoke connectors. I'll take the example of Salesforce. The Salesforce connector we have today in market was purpose-built for Salesforce. It does bidirectional synchronization of data. There's very specific use cases that it solves. And it also has a search before create functionality within Salesforce. So it's very custom built to the Salesforce use case. So what I would say is we treat it as a first class citizen. We are investing in it. We are continuing to build it. But at the same time, Reltio Integration Hub can help you integrate to Salesforce as well. So the way to think about it is, if you have a Salesforce integration use case, first look at the Salesforce connector, look to see what use cases are supported by it.
Ayan Basu (00:22:45):
If it fits that, then use the connector. If it doesn't fit that, and you have some use case that's not currently supported by the connector, then you should look at the Reltio Integration Hub and see if it can help you solve that use case. Because we have had customers that have looked at the Integration Hub and used it for use cases that are not currently supported by the existing SFDC connector. So that's how I would encourage you to think about it. And in that same vein, I would encourage you to think about the Snowflake connector in the same way, where it is, again, a first class citizen and its purpose-build for Snowflake specifically. But then everything else that those connectors don't cover, you're more than welcome to look at Reltio Integration Hub to help you solve it. And then definitely involve us, involve me if needed, and we can help you walk through that and make the best recommendation of which product to use for which use case.
Chris Detzel (00:23:42):
Great. Thank you. If we have five downstream systems, do we need five recipes?
Ayan Basu (00:23:49):
Yes. Each recipe you have a trigger from a specific application and it's a specific destination. So if you have multiple sources or multiple destinations, each one would be a different recipe. Good question. And these recipes can all run simultaneously. So there's no limit on how you want to set your system up.
Chris Detzel (00:24:15):
Can we replace SnapLogic using Reltio Integration Hub?
Ayan Basu (00:24:21):
It depends on how you're using SnapLogic. What a lot of our customers tend to do is they have these iPaaS platforms that are not specific to Reltio. So they're using the iPaaS platform to connect let's say, two random applications. They're using it to connect Salesforce to NetSuite or SAP or something like that. Reltio Integration Hub is design and purpose-built to connect Reltio to something else. So it's not a generic iPaaS platform. It's specific to sources that's using Reltio as either a source of data or a destination. So that's how I would ask you to think about it. So if Reltio is in the picture, if you're trying to get data to it or data from it, then the Integration Hub is something you should use. But if you have two other applications that you're trying to connect, then you would use something like a MuleSoft or a SnapLogic or some other iPaaS platform.
Chris Detzel (00:25:20):
For two way integration using Reltio Integration Hub with Veeva, how much work is required on the Veeva side?
Ayan Basu (00:25:29):
It depends on what you're trying to do. For example, there's a Veeva connector on the Reltio Integration Hub that you can leverage. There's also recipes that we've built. Just a bidirectional synchronization between Veeva and Reltio doesn't require any work per se, on the Veeva side. Where the work might come in is if you're trying to do a search before create functionality on Veeva, much like you would do on Salesforce, then you'd have to think about creating those screens and that logic in Veeva to operate with Reltio Integration Hub. And that's actually... We're in the middle of working on that. Whoever asked the question if there's additional interest or questions, we'd love to be able to work with you, because we're currently figuring that stuff out ourselves.
Chris Detzel (00:26:19):
Great. I put my email address in the chat. So feel free to email me if you want hands-on help from Ayan. How complete is the Reltio Integration Hub connected for Reltio at this point? Any key or obvious missing pieces of functionality? For example, support, partial override support for reverse transcoding of items from Reltio, RDM, et cetera that might mitigate its use in real life scenarios?
Ayan Basu (00:26:53):
No, there's no limitation in the product itself. All the feature functionality is available. It's generally available to everybody. In terms of what you can do specifically with Reltio, RDM or some kind of transformation or however you want to work with Reltio, it really depends on how you build the recipes and what kind of logic you build into the system. But there is no limitation. The product is fully functional and available. And whatever use case you need to solve, you can solve it. And for all the customers that are working on Reltio Integration Hub, I personally haven't come across any limitations. So whatever use cases our customers are thinking of, it appears as though there is a way or a path forward on the Integration Hub.
Chris Detzel (00:27:41):
In a use case of data from a warehouse to Reltio, does Reltio Integration Hub... Is it capable of handling the large volume of that data?
Ayan Basu (00:27:53):
Yeah. The question of volume comes up a lot. This is a cloud native platform on AWS it's because it sits on the cloud and it operates in the same way, like any other cloud software. There is no concern in terms of scalability or limitation in terms of volume of records. So what I would say is, what it comes down to is, what kind of performance do you need? How many records do you have? How do you want to load up the system? You can build recipes to either batch load records into Reltio, or you can do records one at a time. There's an aspect of compute resources on the platform. So we would be able to work with you and recommend the best way to do that. So there's best practices guides and things that we share to help you understand how to build those recipes in the most performant manner. But in terms of just sheer number of volume of records, there's no limitation from that perspective.
Chris Detzel (00:28:59):
They keep coming in, so I'm going to keep asking. Does Reltio Integration Hub have any connection or actionability with DCR, both within Reltio and/or Salesforce.com?
Ayan Basu (00:29:12):
That's something we'd have to explore through the build out of a recipe. We haven't explored that route just yet, but there's no indication that we couldn't get it to work for those use cases. The current Reltio SNDC connector doesn't support it. And that goes back to my earlier comment of, if the connector doesn't support it, then we should explore Reltio Integration Hub. But it seems like it should totally be doable with the platform.
Chris Detzel (00:29:39):
On the volume standpoint, is there performance metrics published from tests?
Ayan Basu (00:29:45):
We haven't published anything on our website just yet, but we do have some metrics that we've shared internally and with our customers. One example that I shared was we saw that we were able to load about 100,000 records with 10 attributes with each record in roughly a five minute timeframe. So we have some baseline metrics like that we can share. And then also, those customer examples that I talked about were in the process of publishing customer references, success stories. And in those stories, we're going to outline performance metrics and details of customer use. So that'll give you a really good idea of what to expect. And then over time as we have some metrics available, we'll be publishing them on our website and documenting those.
Chris Detzel (00:30:30):
Great. Last question for now and we get to the demo. How does it work in case of inbound? So when it comes to real-time with direct source to Reltio versus batch loading with the middle staging area like DF, DW/DL from source to Reltio.
Ayan Basu (00:30:51):
You can use the Reltio connector that's on Reltio Integration Hub, or you can use event queues to pull the queue and see how the messages are coming in. But all of these connections and hookups you have to create within the recipe itself. So on the Integration Hub, you would connect to your data warehouse, wherever you're staging or hosting all your data. You can set up an endpoint to do that, but all of those customizations and even custom connectors, all that capability is available on the platform. So you can build those connection points throughout your integration.
Chris Detzel (00:31:26):
All right. Two more questions, sorry, and then you can do your demo. We find there's no connector using data breaks cluster. Are you planning to add that soon? Or do we have to extend using SDT connectors, like [inaudible 00:31:43]?
Ayan Basu (00:31:42):
A couple answers there. We are building... There was a question earlier. We have a Snowflake connector, we have a Google BigQuery connector. We are also building a Databricks Delta Lake-specific connector to connect to these analytics platforms. But besides that, using the Integration Hub, if there's nothing already off the shelf in the community library on the Integration Hub, you could create a connector to Databricks using the connector SDK, which is a possibility as well. However, once we build that bespoke connector on Databricks, the guidance would be to use that. And then for other use cases, you could certainly look at the Integration Hub.
Chris Detzel (00:32:25):
Yeah. And Bill, you might have to speak up on this particular question, because I'm not really getting it. Bill, can you ask it if you're still there?
Bill (00:32:36):
Yeah, sorry.
Chris Detzel (00:32:37):
Yeah, go ahead.
Bill (00:32:40):
You mentioned the performance metrics aspect of it. And so I assume we would coordinate with our success plan contact to set up calls from Reltio customers that have these performance metrics. Is that right? Is that the channel?
Ayan Basu (00:32:58):
Yeah. There's a couple different ways we could approach the performance metrics. We could either work with you on a POC to actually run records through the system and see real-time what results we get. The other option is the customer success stories and white papers. We're working through that at the moment. So in about three to four weeks, we'll have a white paper published where you can look at that as well. And then the third thing is, as we have some publishable metrics, we will put them on our documentation site, so you get a better idea of what to expect.
Chris Detzel (00:33:37):
Great. Thanks, Ayan. That was a lot.
Ayan Basu (00:33:42):
Very good questions. Keep them coming. And after I show you the product demo, we can talk about what are next steps. How can you get your hands on this and how can you test it and play with it yourself? But before I do that, let me show you and walk you through building a recipe. And in fact, it is as simple as I was saying earlier. And also, I'm going to demo something simple and straightforward for the purposes of time. And I want to be able to showcase to you that it is drag and drop, point and click. So let me just go ahead and do that. So what you're seeing on the screen is just your typical Reltio and the end product. Once the Integration Hub is enabled, you'll see that up here on the right hand side, and you see this Integration Hub icon.
Ayan Basu (00:34:29):
So once you click on it, this is where your whole menu opens up of all the recipes, all the connectors. On the left hand side, you see all the different menus where you can click into. So the first thing I want to show you is community library. And this is the key aspect of this platform, is this is where all of your connectors are available out of the box. And then the other nice thing about this is all of these prebuilt recipes. So if you remember what I was saying is you can use these recipes and customize it. So for instance, if you want to load data from an S3 bucket, you can click "use this recipe" and then essentially just modify and tweak these according to your own use case and needs. But in this case once you set up this recipe, you can pick all the different column names from your file.
Ayan Basu (00:35:26):
As these CSV files are getting into your S3 bucket, you can pick what are the important data points that you're interested in and then set up the actual workflow. And this is going through each of the records that are in the CSV file and then creating an organization within Reltio. And again, these are all pre-built functions that are within the menu. So you're literally just dragging and dropping and I'll show you exactly how that works. So then you go through a mapping exercise and map your different fields to create the organizations within Reltio. And then if you have an error, you can log a message to the output. So all of your error handling, all of your logic, scrolling through, looping through all the records in the file, you can do all of that within this workflow. And so the quick comparison here is this workflow, as opposed to creating a script, writing lines of code.
Ayan Basu (00:36:25):
So it just logically make sense that this would be obviously much quicker for you to put together. So this is one example of a recipe, and we have some other recipes on here too, some recipes around export, and then with queues, there's a recipe. For example, if you use AWS SQS queues within Reltio, as these messages are coming in, you can parse those JSONs and pull that organization from Reltio and then update the account in Salesforce. But it doesn't have to be Salesforce. It could be any other application, just pick your application you want to synchronize to. There's a huge amount of flexibility in terms of what you can do with all this. So that's just a quick glance of what the recipes look like. The way you can think about the community library is it's in something that gets updated on an ongoing manner.
Ayan Basu (00:37:26):
So it's not a static thing that you look at. These recipes are either built by us or our partners build them, or even customers can build and share recipes themselves. So it's really a community of users coming together, much like a developer community. This is a recipe integration community and everybody's sharing what they're building. And so over time, the value that this provides will exponentially increase as the community of users increase here over time. So the connectors that we have, it's about 1,000 right now, but over the next few years, it could easily be 2,000, 3,000. We're just getting started. So there's a lot more to come on this front in terms of available workflows that you can use.
Ayan Basu (00:38:21):
So once you pick a connector, then it's very simple. You just go in here and connect to your desired application, user credentials, passwords, fill out some fields, and you're automatically hooked up. The dashboard is what's giving you real-time insight into how the jobs are working and operating. So you can see failed jobs, successful jobs. And then within the job, you see number of tasks and I'll come back to what tasks actually mean. This is the compute aspect of it, or the actual usage of the specific step or action within the workflow. And so all of those details are visible in a real-time manner, so you can see how that's working. And then finally, on the recipe front, you have different projects that you can set up. So within the projects, you can load up your recipes if you're building something interesting.
Ayan Basu (00:39:19):
So here, there's export, RDM integration, streaming apps. But for the purposes of a demo right now, what I'm going to do is actually build a recipe, a simple one, and show you real quick, how that works. So once you click on "create your recipe" here, I'm just going to type in "test." Select a project, I'll select the app integration project. And then here you have different options. You can either trigger from an app. You can run on a schedule, a web hook, or build an API endpoint. Whatever your desired preference is, you can use that to trigger the workflow and then you just start building. So initially, you want to trigger it from an app. Actually, let me go back and "select trigger from a schedule" so it's going to be easier to demo that. Give me one second. Let me repeat this.
Ayan Basu (00:40:18):
So I'm going to select "run on a schedule" here. And the schedule, pretty self explanatory. Could be minutes, hours, days, monthly, or however you want to trigger these jobs. But then in the action piece, after the next step... And action is where the task comes in, because this is a compute resource. It's trying to do something within the platform. So here, I'm going to select Google Sheets. And just to show you what the Google Sheet looks like, very simple sheet for the purposes of the demo. I just have first name, last name, some names in here to populate into Reltio, which is the use case that I'm going to work on here. And so once I select the app here, what I'm going to do is search through the rows within that file. So I'll click on "search rows" and then connect to the Google Sheets account.
Ayan Basu (00:41:09):
So here, the name of the spreadsheet is RH test. So that's what I'm going to search for in here. So there's a lot of different files here. So I'll click on RH test. And when I select the sheet, I'm just going to select sheet number one, which is what's populated in the file. And then I have a lookup field that helps to actually scroll through the records that are in the file. So once I do that, then I'm going to do a repeat action, because what I want to do is go through those different rows that are in the file and read the values and then put that into Reltio. So here, when I pick the repeat action for the input list, that's just the rows from the Google Sheets. So you can see, all the stuff is autopopulated as I go through the different steps.
Ayan Basu (00:41:59):
And then the next thing is I read the record from the Google Sheet, and then I put that into Reltio. So here in Reltio, I want to create a record and I'm going to connect to the tenant that I'm demoing on. And then here for the object, I'll select "entity" for the object. The object type in this case would be a contact, which is the names that I have in the file. And so once I do that, the last step that I have to do is actually map the fields. So the only two fields I have here are first name and last name. So I just click on these and then apply the changes. And then that should be it for this recipe. So we'll see if this works, hopefully it works. So I'll select "save" and then hit the "test" button. And right now through the "test" button, it's just going into the file, reading the records and populating Reltio.
Ayan Basu (00:42:56):
So you can see the job finished, and I'm going to just go through here and look at the... Actually, I just realized I didn't do one thing that I should have done. So in the mapping, I did not map the attributes in the menu. So if you recall, in "optional fields," I selected first name and last name, and then I actually need to go in here and drag and drop the fields to map them in here. And this is where you can do data transformation if you want. This can get as complex as you need it to. So you can massage the data in however you want and put different rules and formulas on it. But now that I have the mapping in here, now I'm going to retest the job and this time it should function as expected.
Ayan Basu (00:44:01):
So it's essentially going through the records that are in the sheet and populating Reltio. And so once the job finishes, it'll notify me and then we can take a look at the results. So if I go back here in output, this is only picking up the last name. It already went through all the other names. So this entity [inaudible 00:44:25] that was created, I'm just going to copy this and then search for it in the Hub and see if it shows up. Sometimes it takes a little bit of time for it to show up. So I'm just going to put in this entity, [inaudible 00:44:40]. So you can see, Elon Musk showed up as a contact. So that's in a nutshell how it works. And the other thing I want to show you too, is the task usage. So if I go into... Let's see, into jobs, this is the job that ran. And you can see, there were five tasks that were used.
Ayan Basu (00:45:05):
It took about 11 seconds. And the reason it took five tasks is it actually went through the different names. And so on the dashboard, you can also see this job that I just ran. Let's see, I'm going to do just the last hour. This test job ran. I ran two jobs. The first one, obviously I had the missing fields. This one, it's a total of 10 tasks with both jobs combined, but both of them were successful. So you see the result in the output here. So that's essentially it, just wanted to quickly run through a recipe. And you saw how long that took, a few minutes to do that. Again, I'm not saying every single recipe will be that quick or easy, but it's definitely nothing like having to write a script and write lines of code and get something up and running. So I'm going to pause there and see what other questions we have.
Chris Detzel (00:46:04):
I don't know if you completely answered this on the task part, but is there any specific fair usage policy associated with the product which we should be mindful of? So if yes, what is the usage limit that comes out of the box?
Ayan Basu (00:46:18):
Great question. The way we have our contracting process set up today is, whatever number of license profiles you have with Reltio today, and say that number is 100, we automatically give you 25% of your license profiles as tasks. So in that case, you get 25 tasks if you have 100 profiles, but obviously most of our customers have hundreds of thousands, maybe millions of records. So you get a lot of tasks you can run through. Now, if you run over that limit of tasks, say you run over that 25 limit, then a usage-based pricing model kicks in, which is $2,500 for an additional million tasks. But the way we work with you on this is, these recipes, they can be highly optimized and customized. So the task usage, we have best practices where your recipe can work in a way where it doesn't use that many tasks.
Ayan Basu (00:47:18):
So you could batch things up, you could optimize things, you don't have to run it. I just showed you a recipe, which is record by record. But the reality is, for most use cases, it's more a back job, unless you have a need to do record by record. In which case, instead of taking five tasks, you could theoretically just do it in one task. So there's a lot of those best practices that we use. And what we're seeing with the pretty much all of our customers today is that no one is really hitting even close to their task limit. And they're staying well under because they have a fully optimized recipe that's working. So hopefully that gives you some insight into task usage.
Chris Detzel (00:47:58):
What type of transformation is possible? So no Excel-like, but are there any lists?
Ayan Basu (00:48:06):
There are no lists that are published. We have extensive documentation on it. We can share with you a lot of links and guides and things that you can look at, but any transformation you have in mind theoretically, should be possible. With what I've been seeing with our customers is they've been able to do some pretty interesting, complicated things with transformations. The data quality checks that I was talking about earlier, they were checking all the characters within the actual name, checking to see if there's any blank characters, any bad characters, total size. So you can actually put in a bunch of formulas and transformation. Whatever's on this platform is free to use for you. But in terms of what I've seen, I haven't seen any limitations, but we'll also send you the documentation guides. So you can take a look for yourself.
Chris Detzel (00:49:07):
Yeah. And [inaudible 00:49:07]... Go ahead.
Chandra (00:49:07):
Yeah, this is Chandra. Can I do things like complex, regular expression-based transformations, things like that, where we get some bad data? Based on the patent, we should be able to either reject or allow the records, things like that?
Ayan Basu (00:49:28):
Yeah, you can. I believe we have one of our customers that are already doing it in that manner, but we should definitely have a separate session where we can collaborate on what you're looking to do and help you build a recipe to it.
Chandra (00:49:43):
Interesting. Thank you.
Chris Detzel (00:49:46):
And [inaudible 00:49:46] does say, "We have Excel-like functions. However, we also support JS, Ruby, Python-based connectors that allow you to write code, which can be used for building complex transformation and business logic."
Ayan Basu (00:50:00):
Yeah. You can inject your own script, like Chris just mentioned. So a lot of flexibility.
Chris Detzel (00:50:08):
How to schedule a job? How run time failures are reported and can the task be integrated with process tracker?
Ayan Basu (00:50:19):
Yes. You can log your failures to any sort of output. You can integrate it to, let's say Splunk, if you want to shoot errors off into Splunk or your own app profile. Or you could put in error handling within the recipe itself where it stops the job if you run into an error and it sends out an alert or flags an error. So a lot of different options are possible, and customers are using it in all different ways.
Chris Detzel (00:50:45):
Is error handling and retry mechanisms present?
Ayan Basu (00:50:48):
Yes. In the recipe itself, there's a separate error handling section where you can monitor the errors, retry it, stop it. However you want it, you can customize.
Chandra (00:51:02):
And how is the art mechanism? Let's say, I got some data. Sometimes what happens with so many mergers for a period of time. And if something is messed up in the data, I notice that with Reltio sometimes you can only go back to audit only to some extent. So in some cases in the current integration framework, what we have is we have something called staging, keeping the history for some time so that we can go back and cross check where and what went wrong with. If we switch to the Integration Hub, how can we do that? Is there anything on those lines?
Ayan Basu (00:51:48):
Yeah. Actually, let me pull up this recipe that I was showing. You can manually put in messages to an app. This logging message, for example, you can log these messages to an output. And you see this error handling aspect in here, so you can retry. So this is really flexible. You can log every activity that's happening, just errors. However you want to do it. It's up to you.
Garve (00:52:17):
Can I jump in?
Ayan Basu (00:52:18):
Yes, please go ahead.
Garve (00:52:24):
Chandra, I guess you're talking about the job logs and the way to export those job logs. So as part of the Reltio Integration Hub instance that you get, by default, there's a 30 day retention period during which all the job logs are stored. We can increase that for your instance to up to 90 days. However, if you want to keep it for a longer duration, we can configure the instance in a way that those job logs are getting streamed into your data storage, like an S3 bucket for that matter. So that way you can keep a track of all the data integration job logs that you have run for all your recipes. But within the application, we have retention period of 30 days.
Chandra (00:53:13):
Sounds good, [inaudible 00:53:14]. Thank you.
Chris Detzel (00:53:17):
Some other questions. I assume there's some documentation where the best practices for Reltio Integration Hub is recorded.
Ayan Basu (00:53:26):
Yeah. Docs.Reltio.com, you can see Reltio Integration Hub. Click on it and pull up whatever you need to look at. And we're constantly enhancing it and putting up more information there. So if there's something that's missing, something that you want, definitely reach out and we'll keep working on it. And Garve, who's just chimed in, he's working on additional initiatives to enhance our documentation. So it's a work in progress. I can actually pull it up right now, so if you go to [inaudible 00:54:01].
Chris Detzel (00:54:00):
I put [inaudible 00:54:02].
Ayan Basu (00:54:02):
Okay. So you can see all kinds of stuff here.
Chris Detzel (00:54:06):
Cool. For data quality transformations, I'm assuming it can leverage any existing string, cleanser, locate for address, et cetera, that have already been defined.
Ayan Basu (00:54:20):
Yeah. It's interesting you bring up the cleanser. The example that I talked about with our US Census data, that is actually using Reltio Integration Hub, and it's calling out to our external cleanser API that uses locate behind the scenes. So using Reltio where we built a connector to that API, and it's doing all the cleansing and then also appending that data with US Census. That's definitely possible. By the way, what I want to clarify is even if these things aren't available out of the box, you can build all of those things, using the SDKs that we have and create your API endpoints using this interface. So all of that's possible. If there's interest in any specific area, we can have subsequent community webinars where we can answer questions and show you how to do things live like we did today.
Chris Detzel (00:55:20):
Yeah. And we can even work together with other customers to get on the phone and just do some of these shows together. That's the beauty of community is we can help each other in some of this stuff. In case of loading from a file, any limitation on the number of records? So archive mechanisms, is it there?
Ayan Basu (00:55:46):
No limitation on number of records. It depends on performance. Would you want to load that file all at once? Or do you want to split it up? And so it depends on how performant you want your solution and how you want that to look like. But other than that criteria, there's no inherent limitation.
Chandra (00:56:08):
Yeah. I think a lot of people have the similar questions probably. For us, one of the biggest use cases across all the sources and [inaudible 00:56:19] and relations is that how can we handle the quality of the data? We have something called qualified. A record is qualified to get into Reltio versus not qualified, or things like standardizing upfront. Or validating upfront like you said and locate case before even it gets into Reltio. If we want to locate APA to see if the address is verified, it's reaching to the verified level or standardizing the phone number, things like that. Those are really the big things, in my opinion, apart from having a lot of APAs and connectors.
Ayan Basu (00:57:00):
Yeah. And that actually gives us some ideas to build a pre-built recipe on our end to help you with that and get you started. Garve, do you have any thoughts or anything to add on that front?
Garve (00:57:16):
Yeah. I would like to understand that question. Maybe we can have a separate session one on one. Because on the Reltio platform itself, we do support cleansers, whether it's the phone number, the address cleanser or the email cleanser. And you really do not need to go through that cleansing activity before you load that data. Reltio would take care of that.
Chandra (00:57:46):
I agree with that. We can connect one on one and go through more of this. For example, if you have a scenario where I don't want to load addresses that are not verified to Reltio. So I get to know whether it is verified or not after it gets into Reltio and the call happens to the locate and it populates it, not address verification related. All the coding and everything else, status, all that. But before that, what we were thinking about, if we can actually call locate at a certification upfront and load only those that are valid are very fair or reach certain score to be loaded into Reltio. And then the rest of [inaudible 00:58:31] will ping back to our data [inaudible 00:58:34] and have them fix the data where locate cannot standardize them.
Garve (00:58:40):
Yeah. Chandra, I'm not sure if you have seen our new capability around data quality. That's also coming out, I guess today itself, we have a [inaudible 00:58:49]. So data quality dashboards will also point out all those data discrepancies that you might have.
Chandra (00:58:56):
But they come into picture after I load, right?
Guarav (00:59:00):
That's right, yes. There's also efficiency versus performance that we might have looked... But as I mentioned, let's connect one on one.
Ayan Basu (00:59:09):
Yeah. I would suggest, Chandra, if we could have a separate session. Sounds like there's some details to dig into there. Since we only have a minute left, Chris, do you want to help us close this out?
Chris Detzel (00:59:20):
Yeah. And Chandra, if you want to post that question directly on the community, I make sure that these guys get this taken care of. And if we miss any questions, please push them on community.Reltio.com. If you are interested in Reltio Integration Hub and you need some help getting started and everything else, please email me. It's Chris.Detzel@Reltio.com. And I'll make sure that I and the CSM will work with you to make sure we can get you started and get you going. We are serious about really helping you be successful in this and more sessions to come, for sure. Thank you everyone for coming. Thanks so much, Ayan for the demo and the presentation and answering so many questions, which I knew there was going to be. Thanks everyone for coming. This will be posted onto the community here in the next few days. Thank you, everyone.