• AI & Machine Learning
  • Identity Talks
May 15, 2020

SailPoint CMO Juliette Rizkallah on what AI and ML do for identity programs today.

Video Transcript

Hannah Giles: Hello everyone. I’m Hannah Giles and we have Juliette Rizkallah our CMO at SailPoint here to talk a little bit about AI, ML and what that means for identity. Juliette, thanks for joining us.

Juliette Rizkallah: Thank you Hannah, good morning. Good morning, everybody. Well, I think we’re talking a lot about rethink identity at SailPoint and we’re just inviting everybody to rethink their perception of identity and it’s a very interesting topic, the AI and ML. I mean, you know, identities and technology that has existed for several years, if not you know couple of decades and as being through many evolution. But I think the introduction and AI and machine learning in the identity management, especially the identity governance space. It’s just going to be very transformational for enterprise who will embrace that tool. And the reason is, is that, you know, with the explosion of users and the explosion of application that people put in the cloud. Most of the time, this combination of access and users give you an enormous amount of identity data. Identity data that is becoming very, very difficult for human to just sort through and understanding you know how to not just analyze it, but what action take for and from in terms of granting or revoking or changing access. So from that perspective, machine learning and artificial intelligence start building models that will just automate a lot of this right. If you think about big data, AI can just absorb all that data, much faster than any human and start building some learning from it and recommendation. So I really think that this new decade will look at identity and rethink it with all the power of artificial intelligence and how it can really optimize, not just the use of identity, but what companies can do with identity governance.

Natalie Reina: Juliette, a lot of, for a lot of people AI and ML can be very abstract topics, abstract ways of thinking. So can you kind of walk us through what AI and ML and a use case would be for like a real business purpose. Maybe that would help people understand like what it actually means to you to use AI and ML and identity.

Juliette Rizkallah: Yeah, let me try to bring it home. Let’s talk about you know approval request. Right. Somebody wants access and you got to either approve it. You got to decide today. You know, it can be manual. It could be automated. It goes through approvals of managers and so forth. But imagine you having that for a very high number of application, very high number of users, considering that those users can be your employees, your contractors your suppliers, you know, your customers all plugging in into your infrastructure and doing something at some point in there. Everybody’s asking requests and usually, you know, it’s hard for IT who is a little bit more remote from the business to understand why should a person have, you know, access or not. And therefore, it takes time to go back to the manager, get this approval which means the access can take, you know, sometime few hours in the best case but also sometimes few days. With machine learning, they will just look at this access request and look at the peers of that person. Look at what they have, they’re using. Look at the different characteristic of that person based on their role, their job function and so forth. Where they’re located different type of parameters that can be programmed into the machine learning and then will basically automatically said yeah it’s okay to grant, to grant access. This person is in marketing and everybody’s using you know this marketing automation tool. So it’s access, no approval, nothing. You can also be a little bit more control or say the approval is recommended but it still goes to a manager approval or an IT approval. And it’s recommended because usually it’s a non-risky application with non-sensitive data or because in 90% of the case, you know, this was approved. So that hopefully brings it home for you as an example.

Another example on the topic of security, which is also the one that I like the most when I talk about identity governance is that it’s really hard to understand who may have an over provision of access, right? Who may have access to things that they don’t need, they should not have, whether they know they have it or not by the way. Well, artificial intelligence can really build a model where you can see outliers, people that are very different than their peers and have access to a set of application that can be a toxic combination as we as we talk and that graphical representation that artificial intelligence brings just, you know, common alerts, people saying you may want to look at the access profile of this person that is very different of their peer group and that presents some risk. Then there’s an action that can be taken by the IT saying, well, actually that person you know is in IT and needs to have those access for admin. And we’re looking at it closely or could be, oh you know that person should not have access, and then you can revoke the access very easily. Especially in the case of orphan accounts, people don’t even know they have you, that they have access to things. It happens all the time. You and I think, you know, we have access to that set of application. But if you were, you know, going in the back office and see you may find out that you have access to way more application you don’t even know were there. They were just provision to you either when you join the company, when you switch position within the company. A lot of things like that happen all the time and accumulate and it’s hard for a specific IT team to just keep track of all of that you’ve got to have a better way, more efficient way. And identity governance with the power of AI just optimize all that and automated in a way that is, as I mentioned when I started this topic, is going to be transformational for the company.

Hannah Giles: Yeah, that really does that really does hit home. I’m sure for a lot of people. There’s too many apps and systems and data to be managed by human alone. So it’s great to hear, there’s a smarter, better way to do it. And we appreciate you sharing those insights Juliette and we look forward to talking to you again soon.

Juliette Rizkallah: Thank you.

Find out how SailPoint can help your organization.

*required field

By submitting this form, you understand and agree that use of SailPoint’s web site is subject to SailPoint Technologies’ Privacy Statement.