Video transcript
Video transcript
Hi. I'm Rosie, product manager at Tines, and I'm very excited to introduce you to Tines Workbench, our AI powered chat interface.
If you're familiar with or you've tried out other AI powered chat tools in the past, you know that they have a really powerful and impressive general knowledge capability.
Workbench has this general knowledge capability, but on top of that, it's actually able to connect to your proprietary systems in real time, send you real time data about them, and even take actions for you.
And similar to all of our other AI offerings and times, your data is going to be kept within your tenant and it's going to be private. Let's look at an example to see Workbench in action. Say you received an alert that an AWS bucket was made public by a user. You could probably go to any AI chat tool and ask it, what do I do if an AWS bucket let's give it the name too, times demo twenty four, was made public by a user.
And any AI chat tool will probably tell you, the context of what that means and maybe a series of steps that you should then take in that tool, for how to solve this problem. And you'll see that we have that here. It's a potential security risk.
First, we should verify the current status, but Workbench is now going one step further with the ability to access your AWS instance already. So it's it's suggesting we run this AWS CLI command.
Let's go ahead and confirm that. And now it's gonna go ahead and in real time, I don't have to write the CLI command. I don't need to log into my AWS.
It's going to actually access my AWS and get the real time status of this bucket.
It looks like the bucket isn't currently public. Looks like the access control shows that the owner has full control over the bucket, but Workbench is going to check the bucket policy to be sure. Similar to that last time, we can check the CLI command. I'm gonna confirm that, and now it's going to check this. So what this is is actually a template that I've enabled on the left hand side here. These templates are provided out of the box by times, and then you're able to enable them and connect them to your system and even select which actions of these templates you would like Workbench to have access to run or do.
It looks like we've heard back from Workbench. The bucket policy reveals it is indeed public, so it looks like our alert is still relevant and that it recommends we remove the access immediately.
Before I do that, maybe I want to take a look at who this user is. Maybe they had a reason for making it public. So let's switch gears and say, who is Steven O'Brien?
We were just looking at our AWS resource information in real time. Now we're going to be looking into our various systems, to find details about Stephen O'Brien.
This lookup user, you'll notice, doesn't actually have any sort of product associated with it. That's because it's a story. So while the templates are provided out of the box by times, you can build as many custom stories for Workbench to call and run as you'd like. Maybe you have specific custom steps you'd like to take.
Maybe it's more than just an individual action template, and that is, fully supported by Workbench here on the stories tab too. So it looks like we have some frame information from Steven. We checked Okta. We checked Tynes.
We checked Slack for this data. And based on this, it actually looks like his last activity was in August and it's now September. So let's actually check in BambooHR, that's our HR system, if Steven is out of office.
If he hasn't been online since August and it's now September and he's currently making changes to AWS bucket policies, he might that that sounds a little odd. Right? So let's go ahead and look at who's out. It's going to get a list of who's out from BambooHR.
Go ahead and run that for me, and it looks like Steven is in fact out of office for the whole month of September. So this now seems like Steven's account may be potentially compromised or he's on vacation making bucket changes, which probably isn't likely. So I think his account has been compromised.
Can you reset his active sessions and password?
Now we're going to go ahead and take action, reset Steven's account password and active sessions in Okta. Once again, calling a story that I created here and we have confirmation that the password has been generated and reset with some suggested next steps. Yeah. But I actually wanna go back now that we've handled the user, part of this and maybe go ahead and make that change back for the bucket to be private again. Right? So can you make that bucket private?
Now I'm being kind of intentionally vague just to showcase that I don't need to restate the bucket's name. It has full context of our conversation here. So it knows the bucket is times demo twenty four. It knows that it was public from our prime previous messages, and now it's going to remove any public access settings first. Great. Once again, we could view that CLI command like we did previously.
It once it runs that, we're gonna get some data back.
Awesome. Now we're going to remove the policy that allows public access.
Perfect. And once again, you know, I'm being supported by Workbench, but I'm still driving this. I'm deciding that I wanna take this action on my AWS bucket.
You know, there there isn't that potential issue of Workbench going rogue or something and and taking actions that I haven't enabled or allowed it to or confirmed.
Great. Now let's verify that the bucket is private by checking its policy. So we've taken actions, and now Workbench recommends let's go and actually pull the status in real time again, check that those actions did in fact successfully take on our bucket, and once again, we have confirmed there's no bucket policy, that's what we wanted, And then we're gonna check the ACL.
And we have a nice summary of our perfectly private bucket now. So in the span of just a couple of minutes, we've taken a number of actions. We've gotten information in real time about AWS buckets. We've then actually gone ahead and changed to the bucket status and policies themselves and gotten confirmation that those changes have taken place.
We've looked up information on a user. We've also then reset their password and sessions, all within this one chat that has all my context. Maybe I want to give a nice summary. So summarize this info and update the case.
And what was my case? Three seven five two four in a comment.
So I'm gonna update this case comment in times. If you wanted to make a new case in another system or send an email or send a message in Teams or Slack or something, that is also possible. So this is just one example of many for summarizing or updating, but here is our nice summary here summarizing all that we've done. And then now we're going to add that information as a comment to that case. That's perfect. And so it's going to once again call that times template update a case.
Now looking here, let's just check. So this looks like it's gonna update the description. I'm actually gonna say no, make it a comment because I just want comment rather than a full description update. So once again, you can check that and then if you don't confirm, it's not going to do it.
And now we will update it as a comment.
Great. Looks like we're commenting on the case. Let's confirm that and then go ahead and check that out.
And we have this nice security incident update here on the bucket, on Steven's account, on any next steps.
There's so much more we could do and so many more use cases that we could dive into, but I really appreciate you tuning in to just this quick look into Workbench. Thank you so much.