[Music] chris walker your security doctor coming at you live again for another exciting conversation around vulnerability response and predictive intelligence now servicenow's predictive intelligence is a platform function and it provides this layer of ai which gives you more features and capabilities across all of your servicenow applications to just provide a better working experience and so when we think about predictive intelligence with vulnerability response we can start to intelligently classify and build out this framework where now we can take all these requests we can use predictive intelligence to now take specific actions actions that we can train through the solution through machine learning and so we're going to cover two key areas today first and foremost we're going to go through and simply define predictive intelligence and what that entails and then secondly we'll look at machine learning solutions for vulnerability response and showcase that here within a demonstration so let's go ahead and get after it [Music] all right so we're going to talk about predictive intelligence and really what that entails and so i wanted to bring this slide up again because i think it's really relevant but most importantly understanding how we actually train the solution can start to enable your handlers to take all of these incoming requests whether those are incident records or vulnerability items and we can start to lower operational costs by automating the categorization and assignment of those requests right that's gonna reduce task resolution times the number of interactions resolved required to get into those tasks and close them out and of course the error rate right of categorizing and assigning this work really just eliminating the guesswork from trying to understand where these actual operations need to be put in place i like to bring up this slide too because notice when we talk about integrating your vulnerability scanners that's just one component but the automatically prioritized vulnerabilities right the automated triaging predictive intelligence actually plays a key role in identifying and training your solutions through those machine learning frameworks and so by just simply being able to classify these subsets of data the data that we already have we can start to take all of that data and then train the solution to prioritize and assign these particular incidents or vulnerability items to the right teams so they have what they need to make those changes those coordinated plans they can then start to go through and prioritize how they're going to confirm the vulnerability has been resolved so predictive intelligence is extremely powerful and it's an area that you should certainly look into when understanding the framework and how to identify those existing records that have similar values towards a new record and then using that framework to build a word corpus which i'm going to be getting into here just in a moment let's go ahead and actually navigate over to my instance and we can actually start the demonstration all right so we are in my instance we can now see the predictive intelligence home page now as i mentioned before this is where we're going to actually train our solution using machine learning framework so we have four different frameworks here classification similarity clustering and the last one regression now for the purposes of this demonstration we want to focus on vulnerability items and the automatic assignment we want to get the recommendation from predictive intelligence to tell me if i have this vulnerability item and the summary has to do with something specific with a configuration item that's targeted towards that vulnerability i want to be able to then see who i need to actually assign that to in this case the assignment group so let's go ahead and take a look at those solution definitions here i already have this done just to showcase this for the purposes of time i'm going to go ahead and go into the vulnerability item assignment solution now when i go in here you're going to notice that i obviously need to be in the right scope so in this case i'm going to go ahead and edit the record since i'm an admin but it does provide you at least that surface to let you know that you need to be in the right scoped application now that i'm in the editing record for vulnerability response i can see the steps required in order for me to go through and train my solution through using predictive intelligence and then of course using the machine learning components now vulnerability assignment recommendations uses servicenow's predictive intelligence and the solution you're seeing here on my screen to recommend assignment groups for vulnerability items and vulnerability groups now we talked about reducing time that's spent on identifying who this needs to go to for unassigned assigned or even incorrectly assigned vulnerability findings so we can start to see a system generate a confidence score that tells me based on the information you've given me and how you've trained and conditioned the solution 98 percent confidence score this needs to go to the application security team versus the apache team or the network team or the database team or devops teams so now we can start to see who is this best suited for to resolve these vulnerabilities so let's take a look at those steps first and foremost we're going to be looking at step one now step one indicates selecting a word corpus now if you're not familiar with the word corpus this is a framework that would be really good to start building off of now in this instance we're not using one but the word corpus functions as really the vocabulary of this system and how it's used to compare your records based on their kind of textual similarities so for example we can train a subset of records to recommend a resolution based on information from those said records so in this case you can see i have 10 000 plus matching records we're going to use maybe a specific word corpus to help us identify and quickly provide the best resolution for that incident or for assigning out the vulnerability item to that vulnerability group or in this case the assignment group so from there we're going to go ahead and use step two what are we interested in predicting well for the purposes of this demo we want to target the vulnerability items table so we have our vi table already targeted the output is i need to know who this goes to so we're going to train the solution to tell me this is who this should go to based on the vulnerable item the next field step 3 is the input of data what do we actually what's the data we're using to get the output of our predictive solution so we're going to be using the summary like i mentioned before we can take the summary and now we're also going to take the configuration item name we could obviously add more fields to help the solution and train it to better define and detect what type of framework we're going to be using to then give us that output in this case the assignment group and then last what historical data do we want to use from the solution here again we're just using the condition builder so here we've used the assignment group is not empty we have 10 000 records matching our processing language and then of course the training frequency now with this training frequency 30 days is actually a pretty good frequency to start with if you wanted to do it in a much faster frequency you could and it's simply just by going through and toggling down to maybe every seven days so maybe on a weekly basis we want to train the solution to then go through and define and associate those records and then once training is completed it's then going to give us the output in those recommendations but we'll keep it at 30 days because that's a really good subset of data to be able to reference and then of course train the solution from now at the bottom we can see our machine learning solution is active and true we see our solution name the progress of course the row count and all of those other metrics there and then from there we can go ahead and save and update this so we have the ability to update it we have an ability to update and retrain it so it'll kick off the retrain but once that has been completed and done what's going to now happen is we're going to now be able to go into this specific vulnerability item and leveraging that machine learning solution we're now able to identify the recommendation of who this needs to go to in terms of the assignment group so we're actually able to use what's called vulnerability assignment recommendations so we can view that the tool what the system is telling us is the recommendation and who we should assign this to based on a confidence score so because i'm in the vulnerability item here over to the right hand side i have my assignment group and over here i have this little icon it's just a small icon here it's a light bulb that says recommend and when i click on it notice the assignment group recommendations with a confidence score and so i have all these different confidence score ratings but i can see here that my application security team has the highest confidence so they are best suited to address this vulnerability item so now again i'm eliminating the guesswork i'm eliminating hands-on keyboards and trying to understand who this needs to go to who's going to be identified as the right owner of this vulnerability item and other vulnerability items so now i can go ahead and appropriately sign assign this to the application security team now this was just a very high level overview of predictive intelligence and leveraging vulnerability response for a specific use case around assignment recommendations if you'd like more information feel free to schedule a demo with us and as always stay safe [Music] [Music] [Music] you