 Hello, this is Thomas Hazel, founder CTO here at KSsearch, and tonight I'm going to demonstrate a new feature we are offering this quarter called JSON Flex. If you're familiar with JSON data sets, there are wonderful ways to represent information. You know, they're multi-dimensional, they have ability to set up arrays as attributes, but those arrays are really problematic when you need to expand them or flatten them to do any type of elastic search or relational access, particularly when you try to do aggregations. And so the common process is to exclude those arrays or pick and choose that information. But with this new KS Flex capability, our system uniquely can index the data horizontally in a very small and efficient representation. And then with our KS refinery, expand each attribute as you wish vertically so you can do all the basic and natural constructs you would have done if you had, you know, a more straightforward two-dimensional, three-dimensional type representation. So without further ado, I'm going to get into this presentation of JSON Flex. Now in this case, I've already set up the service to point to a particular SRE account that has CloudTrail data, one that is pretty problematic when it comes down to flattening data. And again, if you know CloudTrail, one row can become 10,000 as data gets flattened. So without further ado, let me jump right in. When you first log into the KS Search service, you'll see a tab called Stores. This is the SRE account and I have a variety of buckets. I have a refinery. It's a data refinery. It's just where we create views or lenses into these index streams that you can do analysis that publishes it in Elastic API as an index pattern or relational table in SQL. Now, a particular bucket I have here is a whole bunch of demonstration data sets that we have to show off our capabilities and our offering. In this bucket, I have CloudTrail data and I'm going to create what we call an object group. An object group is an entry point, a filter of which files I want to index that data. Now it can be statically there or a live stream in. These object groups have the ability to say, what type of data do you want to index on? Now through our wizard, you can type in, you know, prefix in this case, I want to type in CloudTrail and you see here I have a whole bunch of CloudTrail. I'm going to choose one file to make it quick and easy, but this particular CloudTrail data will expand and we can show the capability of this horizontal to vertical expansion. So I walk through the wizard. As you can see here, we discovered JSON. This is GZIP file, leave flattening unlimited because we want to be able to expand infinitely. But this case, instead of doing default virtual, I'm going to horizontally represent this formation and this uniquely compresses the data in a way that can be stored efficiently on disk but then expanded in our data refinery on PON query or search request. So I'm going to create this object group. Now I'm going to call this, you know, JSON flex test and I could set up live indexing SQS PubSub, but I'm going to skip that and skip retention and just create it. Once this object group is created, you kind of get to think of this as a virtual bucket because it does filter the data as you can see here when I look at the view. I just see CloudTrail, but within the console, I can say start indexing. Now this is static data. There could be a live stream and we set up workers to index this data, whether it's one file a million files or one terabyte or one petabyte, we index the data. We discover all the schema and as you can see here, we discovered 104 columns. Now what's interesting is that we represent this expansion in a horizontal way. You know, if you know CloudTrail, record 0, record 1, record 2, this can expand pretty dramatically if you fully flatten it, but in this case, we horizontally represented it as the index. So when I go into the data refinery, I can create a view. Now if you know the data refinery of KS search, you can bring multiple data streams together, you can do transformations virtually, you can do correlations, but in this case, I'm just going to take this one particular index stream we call JSON flex and walk through our wizard, we try to simplify everything and select a particular attribute to expand. Now again, we represent this in one row, but if you had to raise and do all the permutations, it could go 1 to 100 to 10,000. We had one JSON object that went from one row to one million rows. Now clearly you don't want to create all those permutations when you're trying to put it into a database. With our unique index technology, you can do it virtually and store it horizontally. So let me just select virtual and walk through the wizard. Now as I mentioned, we do all these different transformations, change schema, I'm going to skip all that and select the order time records event and say create this. I'm going to say, JSON flex view, I can set up caching, I can do a variety of things, I'm going to skip that. And once I create this, it's now available in the Elastic API as an index pattern as well as SQL via our Presto API dialect and you can use Looker, Tableau, etc. But in this case, we go to this Alex tab and we've built in the Kibana open search tooling that is Apache 2.0. And I click on discovery here and I'm going to select that particular view. Again, looks like, oops, looks like an index pattern. And I'm going to choose, let's see here, let's choose 15 years from me from past and present, make sure I find where we're actually was timed. And what you'll see here is, you know, sure it's just one particular data set has a variety of columns. What you see here is unlike that record zero records one, now it's expanded. And so it has been expanded like a vertical flattening that you would traditionally do. If you wanted to do anything that was an elastic or relational construct, you know, to fit into a table format. Now, the advantage of JSON flex, you don't have that stored as a blob and use these proprietary JSON APIs, you can use your native Elastic API or your native SQL tooling to get access naturally without that expense of that explosion or without the complexity of ETL and picking and choosing before you actually put into the database that completes the demonstration of chaos search is new JSON flex capability. If you're interested, come to chaos search.io and set up a free trial. Thank you.