 So I'm going to start off here with an outlier detection improvement which was added to the data quality application in 236. Here I have the data quality application in 236 and you can see that it's changed a little bit from the previous version of the data quality app. I'm going to switch over here now to just show you quickly the data quality app from 235. So this is before the most recent changes and in 235 here we have a bit of a different layout. It's using a less modern tech snack. I'll get into the improvements to kind of the underlying technology and some of these applications that have been introduced to a wide number of applications but you can see that here with some of the slightly improved styling but the functionality changes that have been made in 236 here are we've combined the standard deviation and min max outlier analysis into a single tab. So previously you had separate standard deviation and min max analysis and operations that you could perform to detect outliers in your data values. Now we have just a single now I'm back in 236 with the updated styling and you can see that there's only one tab on the left here for outlier detection. I will select that and you'll note that you have a selection of algorithm here. So we have min max values and z score or z score which is a derivative of standard deviation. So the the functionality is all still there that we had in those two those two tabs previously but this has also now been gives us the capability to expand to additional algorithms for outlier detection such as the ones that Scott outlined when he was demonstrating scatter plots and outlier detection in the data visualizer application. So we will be adding more algorithms to this list in 237 and it's important to note that these outlier detection algorithms are running on the server so they do perform quite well on large databases rather than previously they had to be done in a in a less performant way in the browser. So let's go ahead and demonstrate this. I'm going to select the morbidity dataset where I know there's some data for this particular that with some with some outliers. I then select an org unit and as usual I can select any level of this org unit but I'm going to go ahead and just select Sierra Leone for all of the top level org unit in this instance and you can select a start and end date. You can select the algorithm you want to use. I'm going to start with z score. You can select the thresholds which is the number of standard deviations above the mean that you would you are above or below the mean that you want to detect for this outlier detection and you can also select the maximum number of results that you want to return in this endpoint. There are some advanced options as well such as the start and end date for the data rather than the where they were entered and a sort order as well. So this is going to sort by the absolute deviation from the mean. We're going to go ahead and click start which will generate this report just a moment hopefully. There we go. So now we have a report of the outliers for this particular data set and you can see that there's the z score or z score here in this column. There's also the deviation from the mean which is the absolute value that's the difference between the mean and this particular value and that will give you the sort order for this list as well. Another feature that is in this outlier detection is the ability to mark certain data values for follow-up. So you can mark for instance I will say mark ARI treated without antibiotics and all other new. These are again individual data values that are very high that probably need to be followed up by someone to correct those outliers and we'll see how we get back to that in a moment. If I go now back to this outlier detection I can again select different values here. I could select sorting by the z score rather than the absolute deviation from the mean and this is also going to allow me to determine where that mean is calculated for the deviations to be selected. On the left here we have the tab for follow-up analysis and if we now select the morbidity data set and the parent or unit of Sierra Leone we can select start and end dates here as well and we're just going to leave those as the default. Click those follow-up and we'll see that we have these two data values that I marked for follow-up previously that are now available in this follow-up tab and can be followed up individually. You can then unfollow those if you would like to say that these are okay. They may look like outliers but they're not actually outliers and we can remove those from this list. So that's the first feature here that we have introduced in outlier detection or in the platform set of platform features which is detecting outliers in the data quality application.