Speaker A
I just took and passed the AB900, the agent administration fundamentals exam from Microsoft. And I wanted to make this video to take everything that I learned and break it down all in one video. So then by the end of this,
Griffin Lickfeld, a Microsoft MVP, breaks down the AB-900 Copilot and Agent Administration Fundamentals exam: zero trust, Microsoft security services, and Copilot agent licensing.
Key Takeaways
- Zero trust means never trusting users, devices, or agents by default and always verifying them.
- The three zero trust principles (verify explicitly, least privilege, assume breach) are likely exam questions.
- Success on the exam depends on knowing which Microsoft service feature applies to each scenario.
- Copilot agent administration includes lifecycle management, sharing with teams, and Azure-based billing for Copilot Studio.
What the video covers
- The presenter passed the AB-900 Microsoft 365 Copilot and Agent Administration Fundamentals exam and created a study guide video covering its concepts.
- The video is structured in three sections: security principles, Microsoft services (Defender, Purview, Entra), and Copilot/agents including licensing and billing.
- Exam tips: multiple answers may be correct so pick the most correct one, mark questions for review, manage time (about 45 questions in 45 minutes), and guess since there is no penalty for wrong answers.
- Zero trust is the core concept, built on three principles: verify explicitly, least privilege access, and assume breach.
- Zero trust has six phases, starting with assessing the current security posture, then identity protection.
- Identity topics include conditional access, multifactor authentication, single sign-on, passwordless sign-in, and FIDO2 security keys.
- Access control covers role-based access, security groups, and dynamic groups that add users automatically based on attributes such as location or job title.
- Defender products are covered, including Defender for Identity (on-premises identity signals) and Defender for Cloud Apps.
- Data protection and governance covers data loss prevention (DLP), data security posture management (DSPM), sensitivity labels, and eDiscovery for legal cases.
- The final sections cover agent lifecycle management, the M365 admin center, and licensing, with Copilot Studio billing based on Azure billing policies.
Chapters
- 00:00Introduction and Exam Overview
- 03:00Exam Tips and Format
- 06:24Zero Trust Principles and Phases
- 09:36Identity, Authentication, and Access Control
- 19:00Defender and Security Services
- 33:00Data Protection, Governance, and Purview
- 44:00Copilot Agents and Lifecycle Management
- 55:00Licensing and Billing Policies
Full Transcript — Download SRT & Markdown
Speaker A
you'll be able to understand all of the concepts that you need to to be able to go and take the exam. So let's get into it. Hey, my name is Griffin Lickfeld.
Speaker A
I'm a Microsoft MVP and Power Platform solution architect. And as I said in the intro, have recently taken and passed the AB900. and I wanted to make a study guide video to help explain all of the concepts. And most of what this video is
Speaker A
going to be is going to be a presentation style where we're going to have slides and we're going to work through them. But I will say if you are interested in getting these slides, downloading them for completely free.
Speaker A
The first link down in the description down below is going to be a way that you can download these and join the citizen developer community. Like I said, completely free. You can either follow along or use them as a resource after
Speaker A
you're done watching this. And really my intent in making this is I don't think this exam is that hard if you know what to prepare for. I think there's a lot of things when it comes to agent fundamentals um or agent administration
Speaker A
fundamentals that we can go down the rabbit hole on. And so I want to really get very clear on what it is you actually need to know in order to pass the exam. So the AB900 Microsoft 365 copilot agent administration
Speaker A
fundamentals. How are we going to do this? How are we going to achieve what we need to achieve? So, I've broken this session up into really three main sections. We're going to talk about some of the security principles or more of
Speaker A
the concepts that you just kind of need to know or be aware of. That one is actually going to go pretty quick and then we're going to spend most of our time talking about the Microsoft services that you specifically need to
Speaker A
know. So, this is going to be things like Defender, like Purview, like Entra. And then I want to talk about specifically co-pilot and agents. We'll get into a little bit of the licensing and the billing as well at the end of
Speaker A
this conversation, which are all important. But really where we'll spend the most of our time is talking about the Microsoft services cuz there's a lot of different features all within these different tools that you need to be aware of in order to to pass your exam
Speaker A
cuz you're going to get you're going to get questions on these. So, with that being said, just some quick Microsoft exam tips. If for whatever reason this is your first one, I want to explain maybe what some of that experience is,
Speaker A
but specifically what this looks like in terms of this exam, like most exams, just know that you will probably get questions where there are multiple correct answers, but really ciphering through to understand which one is the most correct is going to be really
Speaker A
important. Also, you're going to be able to I'll speak to an example of when I specifically use this in my experience.
Speaker A
You do have the ability to mark questions as review and come back to them later. So, be sure to do that.
Speaker A
Don't spend too much time on any one question because the reality is this is a pretty short exam. You have 45 minutes and I had around 45 questions. And so it's really about a question per minute.
Speaker A
And so you you kind of have to move fast through the content unfortunately. And so with that, it's really important that you know all these concepts well. And if you feel like you're maybe getting stuck, mark it for review and come back
Speaker A
to it. And then also there's no penalty for incorrect answers. So if you truly do not know, at least put a guess in there. Your best educated guess for the most correct answer. It will not count against you. So that is out of the way.
Speaker A
Let's talk about kind of our first section. Some of these security principles. If we think about what the exam actually kind of looks like, the first kind of third 30 to 35% of our exam is around identifying the core
Speaker A
features and objects of the Microsoft services, which is our next section. Then the next two sections really are talking about understanding data protection and governance. within there is kind of concepts and practical tools.
Speaker A
And so I want to outline some of the concepts first and then we'll really get into more specifically the tools that support these concepts throughout this video. So the first concept that we need to talk about is zero trust. And if you
Speaker A
do the Microsoft learn modules, which I highly recommend you do, this will be like the first thing that you get met with in those modules and I think is the first thing you need to understand. And so this is kind of a concept on the
Speaker A
principle that no user that no device that no application or that no AI agent should be trusted by default regardless of their location. All it just means is that you should never trust anything.
Speaker A
All right? It's zero trust. It is the idea that everything we should assume is a threat, is malicious, and needs to be verified. If we think about zero trust, that's what it is. But Microsoft outlines three specific principles. And
Speaker A
I can I can pretty much guarantee you will be asked, hey, which principle is this? So you need to know these three principles very thoroughly. The first one being verify explicitly. And this essentially means that every time somebody wants to access your data or
Speaker A
log into SharePoint or send a team's message or whatever it may be that we are verifying that this person is the correct person and that they have explicit access in order to do that. And so we'll talk about conditional access,
Speaker A
multiffactor authentication, and single sign on after um this section. We'll talk about that in a little bit, but that's essentially what verify explicitly means. It means every single time somebody is trying to do something in our system, we're verifying that that
Speaker A
person is who they say they are and that they have the access to do that. The next piece is the least privilege access. And so this essentially means that people do not have bloated privileges. And more specifically what I
Speaker A
mean by that is people only have access to do the things they actually need to do as a part of their job. The least possible privileges. Now there may be a scenario where somebody for one specific time once a year or something needs the
Speaker A
ability to do something extra that would not be included in the least privilege access. There's other things that we're going to talk about that will apply to that specific scenario. But the reality is least to privilege access is the idea
Speaker A
that we do not want anybody to have more access than they actually need to do their job. And then assume breach is as it says there the assumption that attackers are either already inside the network or are eventually going to find
Speaker A
their way in. And so everything needs to be protected and assuming malicious intent or that it is already happened.
Speaker A
And so there's tools like Defender, Sentinel, Purview, Entra, things that we're going to talk about on how to mitigate this assumed breach. But these three parts are really fundamental to you being able to pass the exam and they
Speaker A
all make up this zero trust concept. It also outlines kind of six steps or phases of zero trust it calls it in Microsoft learn. I didn't necessarily get questioned specifically on hey what are the six phases? Um these are kind of
Speaker A
more theoretical. Each one of these are kind of covered in more so throughout this video as well. But if we just kind of think through these the first one is hey we need to assess our current security posture. We need to understand
Speaker A
where our company is in terms of zero trust. Then the next thing is enabling identity protection. So think of that as like protecting the people inside of our company, right? We first need to understand where we're at, where do we
Speaker A
stand, then we need to protect everything inside. And then number three, enforce endpoint compliance. So then think of endpoints essentially like your laptop or your computer or tablets or your phones or whatever that may be.
Speaker A
You know, things those are endpoints. We then want to protect all of those things so that you know bad stuff doesn't get in. Once we have everything inside protected and once we're protected from everything outside, we want to classify
Speaker A
and protect data, meaning the stuff that's inside doesn't get outside. And then fifth, and I would almost argue fifth and sixth should maybe even be flipped. But fifth is around monitoring and responding to threats. And then sixth is around educating users. And so
Speaker A
one of the cool things that it actually talks about in the Microsoft Learn module, and I think would likely be on your exam, is that you can actually simulate threats. And so I don't know if you've maybe worked for an organization
Speaker A
where they have sent like fishing emails on purpose that are actually just like fake tests um from them and if you've clicked on them incorrectly it then says like hey you failed this test. Um you can set up those sort of things as well
Speaker A
as you know leading workshops and keynotes around educating users as well. But these are you know if we think about zero trust these are the things we need to be thinking about. All right these are kind of the six phases of how we
Speaker A
implement that. And then we have those three principles as well that we need to be aware of. So that is that is zero trust. I feel confident that you have zero trust mastered at this point. So with that kind of the other concept or
Speaker A
security principle we need to talk about that the exam talks about is passwordless signin. So we obviously are all familiar with signing in and what passwords do. But in today's world and especially in regards to this exam here,
Speaker A
let me actually just go full screen. All right. For the sake of this exam, a regular typical password is no good. If you have a question around just using a regular password, that is not the most correct answer. Any of these other
Speaker A
options around passwordless signin are going to be more more correct in that sense. All right. So, what do we specifically mean is that typical passwords as we know it, you know, username and then a text string of a
Speaker A
password are typically not secure enough for most organizations. And so throughout the um exam you could be asked about these other sign-in types and more specifically around hey which sign-in type achieves this. We have traditional passwords but I have that
Speaker A
xed out. All right that is not what we are looking for in terms of this exam.
Speaker A
If you are in the Microsoft space at all or if your company uses Microsoft Teams and Outlook I'm sure you have Microsoft authenticator on your phone. Right? The multiffactor authentication where it asks you the number you see on your
Speaker A
screen or asks you to enter the six-digit PIN. that is a a a passwordless sign-in type. And I actually specifically got a funny question about Microsoft Authenticator where there was essentially like a screenshot of the authenticator and it
Speaker A
was like what signin type is this? And I was like I actually had to think twice about it. I was like what is this I'm even looking at? This isn't making any sense. I had a question specifically around you know which what is this? This
Speaker A
is Microsoft authenticator. You also have FID2 security keys. You don't necessarily need to know in detail what this necessarily is, but think about it like you have a USB drive. In order to log into your computer, you have to plug
Speaker A
that USB drive into the computer. Like that is that is the thing that allows you to sign in. You have Windows Hello.
Speaker A
This is like the fingerprint or um a lot of them have like face scans now using your webcam when you open up your laptop. That is called Windows Hello. I wasn't necessarily familiar with what that was called previously, but you need
Speaker A
to know that in terms of this exam. The biometrics piece is called Windows Hello. And then certificate-based authentication, which really is just using what are called digital certificates stored on what are called smart cards or devices. You don't need
Speaker A
to know what that is, anything more than that, but those are kind of the different sign-in types that you need to be aware of that I can guarantee you will at least get a couple of questions specifically on. So, it's important that
Speaker A
we understand those. I mentioned that that section would go quick. I want to talk about Microsoft services and this is where we're going to be for a while.
Speaker A
There's a lot of different stuff to cover in terms of of this. So just general M365 services. So you really just need to know that these exist and what they are specifically is you know domain support. So your company's domain
Speaker A
and and what I mean by that is your website right? So your emails are ending with your company name.com or.org or whatever that is and your website is that you can have domain support or domain hosting. You have exchange which
Speaker A
is Outlook. You have teams. You have sharepoint. You have one drive. You have copilot and agents. But you just need to know like that these services generally exist. If you aren't necessarily familiar with one drive, that's really just kind of like your personal storage
Speaker A
that is still in the cloud. Think of it like iCloud if you have an iPhone, right? That that is what one drive is, but just for Microsoft. It's different than Sharepoint and that Sharepoint is group resources. They're shared where
Speaker A
One Drive is for you. It's for one, it's for yourself. You may have questions around the individual admin centers for these. So specifically there is like an exchange admin center, a team's admin center, a shareepoint admin center. You
Speaker A
may be asked specifically around, hey, you're wanting to manage the privileges of this specific SharePoint site. Where should you go to do that? The answer would be like the SharePoint admin center. I think that that's fairly self-explanatory just looking at the
Speaker A
questions or maybe the options that are in there. But a lot of these have their own admin sites and I I wasn't necessarily aware of that before I started preparing for this exam. So just understanding that those things exist
Speaker A
and that are there. Obviously you're not using the SharePoint site to manage your team's chats. That would be the teams admin site, you know, so to speak. With that being said, here are kind of those services. And again, like I you just
Speaker A
need to be aware that these exist. But then when we think about, okay, those are the services. How do we actually monitor those? It's important that we understand these monitoring services because you will absolutely get questions on these specifically around
Speaker A
hey you want to assess your organization's blah blah blah or you want to see how secure your organization generally is or you want to see what steps could be taken to improve the security of your organization like these
Speaker A
things are going to be the answers to those questions so specifically one is called this Microsoft security score this is really just kind of like the general highlevel score that gets assigned to your organiz organization based off of if you have multiffactor
Speaker A
authentication set up, how many passwordless signins do you have set up, what are the security roles look like, how do how many groups do you have, what are the security permissions of those groups, the overall security posture, how many DLP policies do you have, how
Speaker A
many excessive administrative privileges or how many administrators do you have in your system, right? The more administrators you have, the lower your security score would be because there's that many more people who have more access to sensitive information. The
Speaker A
service health dashboard really displays incidents or cases um the exam tends to call them. It's really like if something has necessarily gone wrong or there's some kind of maintenance or anything like that you you're talking about the health of the service health of your
Speaker A
system. Signin reports is really around essentially every time somebody logs in. It's a it's an audit log of this person logged in at this location on this device. So being able to look in if you get questions around authentication
Speaker A
times or when somebody authenticated like think sign-in reports and then the message center which is really around understanding who's sending what sort of messages where those things are going and ultimately providing updates on some new features deprecations and
Speaker A
configuration recommendations in those those sort of pieces. So that's how we we should think about the monitoring services. I also want to talk about when we talk about the services and maybe even monitoring services. One of the things that gets hit pretty hard in the
Speaker A
exam and in the prep is kind of the out ofthe-box builtin protections into M365.
Speaker A
So specifically that is anti-ishing, that is anti-malware, and that is anti- spam. And so you're not necessarily going to be asked in great detail about like, hey, what is fishing? But you really just need to know, like I said,
Speaker A
that Microsoft has builtin protection towards these things. They have advanced algorithms to detect and block these attempts automatically as well as you can set up additional things. I will say being at least aware of what fishing is compared to malware compared to spam
Speaker A
would be important. But nonetheless, I wouldn't say that was like the heavy hitter of the exam. Now, you might wonder like, okay, well, how does it know these things? What what is this advanced algorithm that it's talking about? I didn't specifically get asked
Speaker A
about Mystic or Microsoft Threat Intelligence Center. I don't know if they call it mystic, but that's a pretty cool acronym if it is. But essentially what this is is like a a club or a group. I mean, it's much more than a
Speaker A
club. That's not the right word, but it is a group of people that are actively dark web monitoring and using what's called endpoint telemetry to assess security even at a global scale. So MSTIC is responsible for collecting, analyzing, and disseminating threat
Speaker A
intelligence from a vast array of sources including cloud services, endpoints, and external threat feeds. So like I said, I didn't necessarily get a question specifically around, hey, what is Mystic? But I do think it's important just to understand that like, hey, this
Speaker A
exists cuz it very well could be a question for you. And so this is kind of Microsoft's threat intelligence center that how they are understanding at a global scale, what are these threats?
Speaker A
what trends are we seeing? Are there repeat offenders? You know, those sort of things. If I shift gears a little bit, this is really important before we get into kind of our next section when we talk about security groups and and
Speaker A
and all sorts of different things. But I want to talk about just this concept of role based access control. RBAC, you'll see that acronym if you're not familiar with what role based access control is.
Speaker A
Essentially, what this means is that everybody is assigned a role. Think of it like you're assigned a job title, right? you have some level of job title in what you do. That role or your job title determines what your day-to-day
Speaker A
generally looks like, right? Like what you're responsible for in your current job. And so in the same way, you have a role in some way. And that role is what determines if you can access a specific app or if you can access specific types
Speaker A
of data or if you can modify different types of data or if you can export I'm realizing there's a typo, specific data.
Speaker A
All of this is managed through your role. It's not actually managed like to your email or to your user. Your user is assigned a role. So everybody else with maybe that same job title or everybody else in your department also could have
Speaker A
that same role. The roles actually are what determine your access level. And so specifically roles. There's this thing called users and groups. And this was actually hit on pretty hard, at least in my exam, the difference between these.
Speaker A
And so throughout the Microsoft ecosystem, security access roles, whatever your role may be, whatever your security suite of security you're given or not given is assigned to a role, that role can then be granted to directly to users to specific individual people or
Speaker A
assigned to groups. And what that means is is then you can just assign a specific role to a group and just add members to and from that group. And as they join or leave that group, they get or get that access removed because they
Speaker A
no longer if they're no longer in the group, they no longer have the role. But if they're in the group then they have the role. And so just understanding this idea that there are groups and you have the ability to set groups up. Also
Speaker A
something you know you'll probably get a question around if you're a large organization you should be managing groups and group roles. You should not be managing users and individual user roles cuz if you have hundreds or thousands of employees, it would be
Speaker A
impossible to actually have confidence that every user is set up correctly. Especially if there's users that have you know been at your company for many years and roles have changed and you know requirements have changed and you know all these different things. You
Speaker A
don't want to be managing them on a per user basis. You want to be managing groups. Then you're just managing who's in those groups. Also it talks about there's different group types. So specifically there's kind of two buckets. There's security groups and
Speaker A
then there's M365 groups. Security groups are exactly as they sound. They their main their main benefit is people can join or they can't join but they get added to the group and it gives them a role security permissions, right? And
Speaker A
that's that's really all it kind of does. If they're in this group, they get a role. That's a security group. M365 groups, there's actually kind of some different types of M365 groups that all kind of have different pros and cons.
Speaker A
So, specifically, there's like mail enabled groups. Mail enabled groups can be useful to distribute messages and to grant access permissions to resources specifically in active directory. Then distribution groups really these have no security tied to them. Actually think of
Speaker A
these just as mailing lists. This is part of the exam part of the module learning. You will probably get a question around hey you want to use groups to manage users. Which type of group can you not use or manage users
Speaker A
permissions? Which sort of group can you not use? You cannot use distribution groups. All right? Because distribution groups are simply just like mailing lists, right? So if you want to be able to email a group of people, but that
Speaker A
group of people has no real sense of a unified role or security, you want to use a distribution group. The dynamic group is a little bit different though.
Speaker A
And what that means is this you can add roles to. But the difference here is dynamic groups is that people automatically get added or removed from the group by the system. So how that works is you can actually set up like
Speaker A
filtering criteria for users. And so like typical security groups, there's an admin that has to add or remove users from the group. Where a dynamic group could be used and say, "Hey, everybody in the state of California with the job
Speaker A
title of sales executive. If they meet that criteria, they automatically go into the dynamic group for that. No user, no admin has to go and add them to the California sales exec group. they just get added instantly based off if
Speaker A
they meet the filtering criteria and roles can be added to dynamic groups which then be inherited by all the people that are getting dynamically added to the group. So those are groups those are kind of the different types of
Speaker A
groups in there and how we need to be thinking of role-based access controls and how we can be applying them to users and to groups. So I want to talk about the first of one of our security services I guess I would call them and
Speaker A
that is Microsoft Defender and understanding what Microsoft Defender truly is. Microsoft Defender or Microsoft Defender XDR is a comprehensive security solution designed to protect organizations from both known and emerging threats. So what does that actually practically mean? So Microsoft
Speaker A
Defender practically is a security tool that helps protect your computers, your phones, and your data from external threats. And like if I could super oversimplify it, if you see Defender, think it keeps bad stuff from getting in. All right? Like that's that is that
Speaker A
is what defender is. It keeps bad stuff out from being able to get into our system. Some of the other services that we're talking about will be more focused on you know maybe keeping stuff in and not getting out. But like if you think
Speaker A
about defender it's hey we're trying to keep bad stuff from getting in. We are trying to defend our system with that.
Speaker A
One of the things you will see on this slide is that there are both pre and postbach capabilities of defender. So this is not just like a you just set up the rules and hope it works, but Defender also helps you actually see
Speaker A
advanced analytics for threat intelligence to block malicious activity before it actually reaches any users.
Speaker A
But then also it has really cool investigative and response tools inside of it in the event that something actually does happen, if the event it something does actually get in. And actually inside of Defender, there's kind of like four children or four
Speaker A
pieces of Defender. And I wasn't necessarily tested on the differences between these extensively, but I think I think they're fairly intuitive just simply based off of the naming.
Speaker A
Specifically, we have Defender for Office 365. This is going to protect our email. This is going to protect our collaboration tools like Teams and SharePoint. It's going to protect our files. It's going to protect us from fishing, from malware in our emails.
Speaker A
It's going to have the safe links and the safe attachments and the anti- fishing policies, right? Everything in terms of Office, Outlook, Teams, SharePoint like Defender for Office 365, Defender for Endpoint is really around devices or what we kind of call using
Speaker A
behavioral analytics and kind of ultimately kind of detect and respond to threats. This is not a technical exam.
Speaker A
So, you don't necessarily need to know the the nuances of like different types of endpoints and networks and you know the perks of VPNs and like all of these this stuff. Uh you do you are not tested on that at all. But just like if you see
Speaker A
the word endpoint and it's in relation to keeping bad stuff out. Defender for endpoint. Defender for identity is really it's really not like intra ID or like user identity. It's really like onrem active directory stuff. I didn't have any sort of questions on this but
Speaker A
if you see something around on premise identity or you know user authentication on premise authentication or on premise signin or anything like that defender for identity and then you have defender for cloud apps so just think of this as
Speaker A
you know power apps or dynamics any other sort of thirdparty apps or code apps that you've added to your system there could be threats that are coming in you have defender for cloud apps and so that's that's really defender I think
Speaker A
it's fairly self-explanatory Right? You're trying to keep bad stuff out. But if we move then to Microsoft Sentinel, Microsoft Sentinel is really the visibility tool. All right, this is essentially a log of alerts across M365 and even non-Microsoft services or
Speaker A
sources, excuse me, that you add. And so specifically what this means is this is kind of like the central security command center for your company. What that specifically means is that this is the tool to help you detect,
Speaker A
investigate, and respond to cyber threats. again across your entire system. So, inside of M365 and outside of M365. So, like we said, Defender is kind of the more active, hey, here's how we protect bad stuff from getting in.
Speaker A
Sentinel is like the the visibility into just the massive log of any potentially detected threats. Like we said, there are some visibility pieces inside of Defender. Sure, not saying that there aren't, but in terms of the exam, if
Speaker A
you're looking for if you're getting some level of question around the log of everything that's happening inside of your organization and could even including a third party service, you should be thinking Microsoft Sentinel.
Speaker A
The next service we're going to briefly touch on, which Sentinel is is pretty wasn't really hit on very heavily even in the modules or the learning materials, but now if we talk about in tune, this also was kind of a smaller
Speaker A
one. If you see in tune, think compliance. All right. So, specifically what we mean is, and you know, I got to hand it to the Microsoft naming here as far as like something is in tune. Um, if I'm not a music person or I I don't know
Speaker A
how to play music or anything like that. I don't know how to play any instruments, but I love music. But with that, if it's it's in tune, it's compliant like with the song or whatever. I that's how in my brain I
Speaker A
have pictured it. It does much more than just simply compliance. But from like an exam perspective, the differentiator was the word compliance. There there are some other pieces on here and that it can help you protect your data. Protect
Speaker A
the data that the people at your organization access. Safeguards the devices. Ensure the organization is using proper credentials. Ensure that all corporate services are available to end users. Help new members of your organizations have a good onboarding experience. Ensure that users get a
Speaker A
first-class support experience. There are some other pieces to this. But again, for the sake of the exam, if you see compliance, generally think in tune.
Speaker A
There there's one wrinkle to that that we will talk about in a second which I guess I'll spoil it is the compliance manager inside of purview but that is that is very clearly different as well than in tune. So if you're thinking hey
Speaker A
what Microsoft service should I generally be using to you know in regards to compliance think in tune.
Speaker A
Let's talk about perview and I got to be honest perview is the big dog on the block where there are lots of features that are super important to the exam that all kind of fall under purview. So, if we think about, you know, we're kind
Speaker A
of we're talking about M365 services. Underneath that is perview, and underneath that is nine key features that we're actually going to talk about.
Speaker A
And you kind of need to know the nuances of each of those nine key features.
Speaker A
Hopefully, this isn't feeling overwhelming. You got this. But I also think that a lot of times the naming of these can can um really help clue us into what it is they do. So, hoping I can bridge that gap for you. Let's read
Speaker A
this. Purview provides a centralized suite of tools to help organizations manage their data across Microsoft 365 and other cloud or hybrid environments.
Speaker A
Let me just go over these. So here are kind of the nine key features that we are going to go through fairly briefly that all fall under purview in some way.
Speaker A
If I were to summarize these, you know how we said defender is like keeping bad stuff from getting in. Think of purview as keeping inside stuff from getting out. If defender is keeping bad stuff from getting in, like it's blocking bad
Speaker A
stuff from getting in. Purview is like the keeping inside stuff from getting out, if that makes sense, right? If I if I could use my hand gestures, that's that's how we should be thinking about this. And I realize my face is kind of
Speaker A
covering this one. Content search and eiscocovery we will get into as well. Let's go through kind of these first three key features. The first one is information protection. And so really what this means is labeling. And so this
Speaker A
means you have the ability to label documents or any sort of data as highly sensitive, not sensitive and anywhere in between. What could be a sensitive document or a sensitive piece of data.
Speaker A
Think about if you like an onboarding document or a W2 form or something, right? Like where there is maybe bank details or a personal identifiable number in the United States, social social security number. If the document contains sensitive information, it can
Speaker A
be labeled as sensitive and this kind of falls under information protection. One cool thing about this is labeling can be done both manually and automatically. So there are machine learning algorithms that are built to automatically detect and label documents and the modules
Speaker A
specifically talk about using you know social security number as an example. So specifically if it detects these sort of information or this sort of information it will automatically label it as highly sensitive. And so just that concept of
Speaker A
labeling is really important to know. What that means though is then how we kind of like realize labeling is with data loss prevention policies or DLP policies. And so this essentially is what actually then acts on labeling.
Speaker A
Imagine somebody is trying to share a document whether it's in an email or in a team's chat or something. If I add that attachment to my email and then I go to send that email to somebody and that document happens to be labeled as
Speaker A
highly sensitive, purview will block my email from being sent whether intentionally or unintentionally, right?
Speaker A
Determine how the admin sets up the DLP policies, right? It will block the ability to share information. So data loss prevention all right it's really thinking about hey how do we make sure that data not that data doesn't get lost
Speaker A
in our system but like that data doesn't spill out of our system all right or spill to other departments still inside of our system right that like stuff isn't going to the wrong people all right and again it's it is kind of built
Speaker A
off of labeling insider risk management I think this one's pretty self-explanatory based off of the name think of this as bad actors in your company so specifically insider risk management allows you to the risky behavior that people inside of your
Speaker A
organization may be doing. This could be weird signins. This could be weird downloads. This could be weird size of downloads. This could be weird exports, imports, whatever it may be. You likely will get some level of question around,
Speaker A
hey, you're suspecting that an employee may be malicious in some way, right? like the tool you should be looking to leverage in that scenario inside of risk management inside of purview or maybe like another example is hey you were
Speaker A
alerted that an employee that last night downloaded 10 gigabytes of data onto a flash drive or something that alert and how you mitigate that or how you see that is done in insider risk management like that's the answer to that question.
Speaker A
Those are our first three. Our next three and again there's nine. Communications compliance honesty it's exactly as it sounds. So let's let's try to keep this stuff simple cuz there is a lot of it. This manages monitors and
Speaker A
manages communications. It's across teams, email. This continues to ensure that communication is inherent to regulatory or organizational policies defined in communications compliance. I want to spend a second talking about data security posture management or DSPM. Specifically, what this is is this
Speaker A
is a tool that makes sure your data is safe and not overexposed again inside of Purview. And so really what this is is like a visibility tool. We have DSPM, but then we also have this thing called DSPM for AI, which you probably will get
Speaker A
tested on that alternative as opposed to just typical DSPM, but more specifically DSPM for AI is an extension of data security posture management that's focused on protecting the data used by agents or in AI systems. And so in
Speaker A
simple terms, it makes sure that AI is using safe, secure, and appropriate data. And in all honesty, DSPM was a little confusing to me for a couple of reasons, but really what it means is that this is both a concept and also
Speaker A
something you kind of actively configure in DSPM in Purview. And so what this specifically means is yes, it is kind of the concept in terms of being able to look at it, being able to understand our general security posture all the way
Speaker A
down to like security rules and Microsoft Dynamics or Power Apps like our security posture and security posture with AI just like our Microsoft security score, right? But more specifically, um you can actually connect different data sources including external ones and in DSPM actually
Speaker A
determine who or what agents can access what data. So, say you want to block a certain data source from all agents to be able to use in your system, but you still want people to be able to go and
Speaker A
access it, but you want to be like make sure no agents are ever able to touch this system or this table or whatever it may be. You do that with DSPM for AI.
Speaker A
One final thing on that, make sure you know what DSPM stands for. I I had a question that was specifically around like what is DSPM? And in my studying, I hadn't really prepared enough to memorize the acronym. And so I was like,
Speaker A
DSPM, what is DSPM mean? Cuz I think I think if you saw data security, posture management for AI, it makes more sense, right? Like the name makes it more obvious for you. But again, I this question only shared the acronym. And so
Speaker A
what I did was I actually just I marked it for review and I was like maybe there's going to be a question later in the exam that says the full name or you know, my memory gets jogged on what it
Speaker A
actually means. And that did happen to me actually. Um like probably like nine questions later I got another question on data security posture management for AI DSPM in parenthesis and I was like okay that's what it is and I went back
Speaker A
and confident that I got the the question right but I would just say you might not have that the roll of the dice. I guess I got lucky in that scenario. So little tidbit pro tip make sure you at least understand what that
Speaker A
DSPM means data security posture management. Final piece here. I'm realizing I'm kind of covering this. So, let me flip to the other side and cover the photos. So, specifically data life cycle management. This is as it sounds, right? This is how data is created. So,
Speaker A
that's like the birth of the life cycle so to speak, right? Like how it's created, how it's then updated, how it's then stored, how it's then archived, and ultimately how it is deleted. And so that process if we're thinking about
Speaker A
storing data over a long time that would fall under data life cycle management of Microsoft purview. All right. So if we think about use the word life cycle from creation to removal or deletion anything in that being able to manage when that
Speaker A
happens maybe automatically or how long we hold on to stuff after it's hasn't been used. Right? So hey this contact hasn't been touched or been on a new opportunity in the last 4 years. let's archive it and then if it remains
Speaker A
archived for another 3 years we delete it or something like that like that process would be under data life cycle management. The final three is compliance manager, data explorer and then content search and eiscocovery. So compliance manager this is actually
Speaker A
something that I had seen for the first time at Microsoft Ignite um in the fall of 2025. And so specifically the compliance manager, think of it like the Microsoft security score where it's kind of just a it's taking a pulse of your
Speaker A
security posture in your organization and it's saying you're generally this secure, giving you a grade. Basically, the compliance manager is pretty similar just in terms of compliance. All right, it's that same kind of concept. So it allows you to kind of compare yourself
Speaker A
to different compliance global standards or regulations and it actually allows you to not just see a score but actually identify the gaps and act and uh act on those gaps. So it would can it will actually kind of give you little tasks
Speaker A
like hey complete this and you'll be thus two more points compliant or something like that. That's the compliance manager. Things around, hey, we want to see how compliant we are in regards to HIPPA or we want to see how
Speaker A
compliant we are in regards to our governmental regulations or we want to improve our compliance, but we're not sure where to start. Think compliance manager inside of Purview. The next piece is data explorer. And so really what data explorer is is a way for you
Speaker A
to see where sensitive data exists in your system. data explorer. This allows you to see data classifications and labels kind of like what we were just talking about um and labeling and even like drill down into the actual files
Speaker A
and the emails dependent on your permissions and everything like that. But ultimately really what data explorer is for is you want to know hey where where are people putting credit card numbers in my system? Where are people saving social security numbers in my
Speaker A
system? Where are people doing this? Where are people doing that? Data Explorer shows you, you know, the files in SharePoint with the credit card data or with the emails containing that info or how many items and where those items
Speaker A
are inside of your system. All right, so it's it really is the explore. It's the way to search and kind of see where sensitive information is. Now, how is that different than content search and eiscocovery? Content search and
Speaker A
eiscocovery, think legal investigation or case resolution. The exam and the modules call them cases, right? So, think about you're in a legal investigation. That is what content search and eiscocovery is for where this allows you to search through emails,
Speaker A
search through files, search through teams messages, search through whatever to kind of build your case or to find evidence so to speak. All right, so specifically you can use keywords, you can use search for users, you can search
Speaker A
for specific dates and ultimately let you export those results. So what do I specifically mean? Data explorer like I said, hey, we want to see where people are saving this sort of information in our system because this would be
Speaker A
sensitive. We want to make sure it's managed in the places we want it to be managed. Eiscovery is like, hey, our company is in a lawsuit. Our legal team has now created a case. We need to search all relevant emails and files and
Speaker A
that could potentially apply to this case as evidence, lock them so they can't be changed, and ultimately review and export them for when we take them to court. Like that, that is the idea of what eiscocovery is compared to data
Speaker A
explorer. I'm hoping that distinction kind of makes sense. That wraps up kind of these nine key features inside of Purview. The last service that I want to talk about and I realize I'm still I'm not covering this is Microsoft Entra.
Speaker A
All right. So, Entra at the end of the day, think of it as, you know, this is built on role-based access control and zero trust. But Entra is where you are actually managing your people's identity, your users in the system.
Speaker A
There's a couple of concepts or even more than just concepts that are managed inside of Entress. So, just like how perview had nine key features, Entra kind of has, it's like two and a half really. I know there's three on the
Speaker A
screen, but I'll explain here in a second. So, the first one is conditional access. So, specifically, this allows administrators to define roles based on users, device, location, and risk level.
Speaker A
And so, specifically, you may have your user may have access to see certain data, but conditional access and let's say you live I used the state of California earlier. Let's say you live in California, but then for some reason
Speaker A
at 2:00 a.m. Pacific Standard Time, somebody in Europe logged into your user and is trying to access this information. The likelihood that that is actually you is very low. And so conditional access is actually looking at the signals based off of your trends
Speaker A
to see if there are any risky behaviors. Even though your credentials, you know, they logged in successfully, even though your credentials have access to do that, conditional access, which is managed in Entra, is actually managing risk signals. And so you can set up these
Speaker A
risk signal policies underneath conditional access inside of Microsoft Intra. The other piece I want to talk about is privileged identity management.
Speaker A
And so privileged identity management is where you can enable just in time permissions. So what I mean by that is earlier I mentioned when we were talking about least privilege earlier I mentioned like say there's one time a
Speaker A
year you need the ability to see some extra information for whatever reason. Um but it's it's very it's not really a part of your typical job. If we were considering least privilege it's definitely not you know a part of that
Speaker A
you can achieve giving you additional permissions with privileged identity management in Microsoft Entra. So you can actually set like time frame the person and what they actually going to get access to and when that time comes they lose those permissions. So you can
Speaker A
set up what's called just in time permissions. Now just in time permissions are really more of a concept and the pim is really more how you kind of implement it or actually exercise it inside of the Microsoft ecosystem. But
Speaker A
just in time permissions is exactly what I said it just was. So it gives you additional access to a specific group of people for a specific time or through a specific channel or medium and ultimately once that task is complete or
Speaker A
the time is up the elevated privileges are then automatically revoked exercising the least privilege principle of zero trust. That's kind of what we want to think about as far as Microsoft intra. And so if you're getting questions around we need a user to be
Speaker A
able to do this. I keep using the example once a year but they have some additional requirement where they need this thing but it's not a part of their typical day job. You should be looking at PIM or JIT or underneath entra. So
Speaker A
let's put this all together. All right. I want to kind of put it into one slide to kind of compare all of these and how we can think about these. So we have our M365 admin center the centralized portal
Speaker A
to manage our services, manage our users, manage our licenses and any major settings like the domain across our system. We then have Defender, which is the full suite of security solutions to keep bad stuff out. We then have
Speaker A
Sentinel, which is the ability to collect and analyze and ultimately respond to data security threats across our entire organization, both inside Microsoft services and outside. We then have Intune, right, which is the cloud-based service for managing compliance across devices, apps, and
Speaker A
security policies. We then have Purview. Purview is how we keep our good stuff in. It's the solution for discovering, protecting, and managing data from escaping the organization. Then we have Entra, which is the family of identity and access solutions, including Azure
Speaker A
Active Directory. It's not inside of the admin center, but when you're looking at users in the admin center, you're actually looking at Entra. Entra is what's actually managing the users, the identity, the conditional, the role-based access, the conditional
Speaker A
access, the just in time or the privilege identity management policies, and all of those things. I think that actually controls which people have access to which resources. That is the end of section two. Like I said, that was the huge bulk of this conversation.
Speaker A
I challenge you take two seconds like me to take a second and take a break. Let's talk about co-pilot and agent administration. So specifically, if you've been on LinkedIn or in the Microsoft industry, you have probably seen this graphic back when Microsoft
Speaker A
Copilot came out. Everybody was posting this thing and would instantly get a bunch of reactions cuz it looked so fancy and looked super cool. But why do I have this here? Copilot and agents built in Copilot Studio are connected to
Speaker A
a lot of different things, right? They're connected to the users and the M365 apps in the top left where people are talking to them. They're connected to Microsoft Graph or any sort of services that you connect to your custom
Speaker A
agents, the flows and all the actions. And then they're connected to the models as well. But what you can really see as a part of this is that everything is going being filtered through the middle.
Speaker A
All right, this co-pilot icon. And while it does look cool like that, what that's actually signifying is that Copilot is an orchestrator. And so specifically what that means, there are pros and cons, strengths and weaknesses, and even certain limitations due to the fact that
Speaker A
we're using C-Pilot. And so what do I specifically mean? What I mean is what the user sends. Say I'm in Word and I send something to Copilot, it's getting processed or filtered maybe would be the better word, by Copilot. If it grants it
Speaker A
like safe, then it sends it to graph. And then if it grants that as safe, then it sends it to the LLM. And if it grants what it came back as safe, then it sends it back here for grounding. And
Speaker A
everything is being filtered through co-pilot or copilot studio as our orchestrator. Let me maybe say a practical example. So I had it wasn't even it wasn't me. I had a friend um come up to me and share with me a use
Speaker A
case that they had been asked about. And the use case was essentially an agent to help us track criminals. And specifically, this agent would need to be able to reference criminal histories, which obviously can be criminal histories, right? They can have very
Speaker A
sensitive information in terms of what the crimes were, maybe graphic or unfit for children to read, so to speak. Maybe that's the best way to say it. Co-pilot will not talk about crime. It will not elaborate on murder or something like
Speaker A
that, right? It will it'll see that and be like, "Whoa, I'm not having this conversation." All right? like that is a limitation that copilot has from Microsoft and there's nothing you can do to get around it. I'm not saying these
Speaker A
limitations are bad. All I'm saying is that it exists, right? Because everything is being filtered through this orchestrator. There's kind of some steps and it kind of aligns to what we were just showing there around these co-pilot orchestration steps. And again,
Speaker A
I wasn't necessarily tested specifically on this order, but if we're thinking about, hey, we're going to be agent administrators, which is what this exam is about. We need to understand what the orchestrator is, where the orchestrator is affecting what we do, and ultimately
Speaker A
affecting our agent and the end user. The first is kind of the and I I do think this makes sense like if you just kind of read through it and it's it's fairly self-explanatory, right? The prompt intake, the orchestrator is
Speaker A
filtering what the prompt was sent cuz the prompt could be sent from an intended user or it could be sent from a malicious user or somebody who doesn't actually have access to that agent or they're trying to prompt inject extra
Speaker A
information like that prompt intake is going to be filtered through the orchestrator. Then it has what are called evaluations which is where it's like digesting that message and granting it safe. Then it takes it to grounding.
Speaker A
That's where it went to that bottom left and performing what's called rag, whether that's on Microsoft Graph or your knowledge sources or whatever that may be for your custom agent. It's adding knowledge and business data to the users prompt that you sent. It's
Speaker A
then passing it back over to the LLM. We saw that going through. Then once it comes back from the LLM, it's going back to kind of regground it. That's that formatting where the agent's cleaning it up, reggrounding, and ultimately
Speaker A
producing the readable output from the LLM, right, with the business context, then filtering that back to the user.
Speaker A
So, I'll go back a slide. If we apply that back to this graphic, right, we even have the steps here. 1 2 3 4 5 6 it's following this trend. That is what is happening. So that's copilot as an
Speaker A
orchestrator. I want to talk about then something that the exam talks about a lot is there's different platforms so to speak to be using co-pilot or to be building agents or using pre-built agents in the system and you just need
Speaker A
to know that these exist and maybe some of the differences or distinctions between each of them. And so that's what I want to do now is how do we know like we're selecting the right tools and the different tools that are available to
Speaker A
us. And so the first one we have M365 copilot right so M365 copilot think of this as your general one-stop shop to talk to anything regarding outlook to teams to one drive to share to Word Excel PowerPoint it's the one-stop shop.
Speaker A
This is the general productivity tool. Think about this. If a business user needs to quickly summarize a report, analyze last month's sales figures, co-pilot is the right choice. So ultimately think about this where IT admins should see co-pilot as the
Speaker A
frontline tool for enabling productivity. Hey, we want to boost productivity. Think M365 copilot. Then what we should be thinking about is pre-built agents. So pre-built agents, these are things like researcher and analyst. If you've taken the AB731 or
Speaker A
the AB730, you're more than familiar with these. IT admins should consider pre-built agents. These are things not built by your organization but like built by Microsoft pre-built agents whenever the volume of requests is predictable and workflow is standardized
Speaker A
or you know what we're seeing is a lot of people essentially doing research researcher type tasks with M365 copilot then we should potentially look to leverage pre-built agents. We then have agents for business users. Agents for business users. Think of this as the
Speaker A
agents you can build inside of M365 Copilot. I don't even know what it's called nowadays, but you know, copilot studio light or agent builder, whatever that experience is, right? This is not co-pilot studio, but when you can build
Speaker A
those simple business user focused agents inside of M365 and share them with your team members, that's what we're ultimately talking about here. So, IT admins should encourage this approach when the solution doesn't require any sort of external connectors or what what
Speaker A
I have here is system integrations. um it doesn't need to be accessible to users outside of the tenant or outside of the environment. But ultimately these are suited for like little lightweight department specific needs. Not necessarily M365 C-pilot, but you want
Speaker A
it to be kind of the same thing, but maybe more targeted to something more specific. We should be looking at agents for business users. Then we have SharePoint agents. Pretty self-explanatory that this is going to be accessible inside of SharePoint and
Speaker A
we wanted to be able to look at content stored in SharePoint inside of that site. Then we should be looking at SharePoint agents. Then ultimately the next iteration and probably would be unfair. There's probably another one even to the right of this. And I realize
Speaker A
I'm now covering this. There's probably even one more to the right of this for Foundry, but that is not remotely close to the content on the exam. Custom agents, we're talking about when the workflow is maybe complex, when it needs
Speaker A
to connect to multiple systems, when it involves maybe automating tasks or connecting to agent flows, requires organization specific logic, etc., etc.
Speaker A
So that's cool. We understand the tools, but like how do we actually manage our agents after we build them? We talked about data life cycle management a little bit, right? When we were talking about purview, well, we also just need
Speaker A
to understand agent life cycle management. This is another kind of piece of the puzzle in terms of this exam. And so just like data life cycle management, agent life cycle management is the comprehensive framework for managing AI agents or automated systems
Speaker A
throughout their entire existence. I'm pretty sure I copy and pasted this directly from the Microsoft learn module. So, let me just read this. Says, "Unlike traditional software, AI agents are continuously learning and evolving, which requires ongoing oversight to
Speaker A
prevent errors, security risks, or performance drift. An ALM ensures that agents are treated as longive products, not one-off features that we're creating accountability and operational reliability." So especially as we think as models changing or new models coming out or our business process changing the
Speaker A
process for hey our business process is now changed or how we want this output to look has now changed or anthropic just came out with a new model that looks like it would be really great for us to use like how we go about updating
Speaker A
our agent making sure it's continued to be reflective and correct is ALM or agent life cycle management kind of outlines five steps of ALM I think much like the other steps that we've shared throughout This session follows a very
Speaker A
similar structure, right? We have creation. I mean, you could probably put a 1B as in like build, right? But then two, we have approval and deployment, which is, hey, once it's actually built, it's tested, it's verified that it's
Speaker A
connected and working correctly, we then actually publish it, make it available to users. Then there's maintenance, which is over time. Again, as we say, things change, models change, processes change, documents change, all data changes, all these different things, the
Speaker A
maintenance, and then ultimately into the block or thus removal of agents. So you have the ability to block agents if it seems like it's not operating with the amount of confidence that you wanted to operate with or there's some level of
Speaker A
threat or it was surfacing sensitive information to the wrong user. You can instantly block it. That is very different than removing it. Blocking it, you can fix it and then reenable it, but then ultimately bringing it to like a
Speaker A
removal. So think of block as like the archive of data and removal is like the delete as in the agent might no longer serve its purpose. Business has changed so much that we're just going to do a new agent. That sort of thing. that is
Speaker A
agent life cycle management. Kind of the last section for us to talk about is around licensing and billing policies.
Speaker A
And a lot of this licensing stuff will probably feel pretty similar if you have the AB730 or the AB731.
Speaker A
This exam doesn't even talk about tokens, but it's really important for you to understand what tokens are for the rest of this stuff. So tokens are essentially the foundational unit of AI processing. So every time you ask co-pilot or an agent or chat GBT a
Speaker A
message, every time you send it, it's using tokens as the the currency or the unit of measure for how it's tracking how much energy it's using to to process that information. So that's kind of the foundation and those tokens ultimately
Speaker A
get charged to organizations or you know people or organizations consuming those tokens. There's kind of a couple different licensing types that you need to be aware of. There's subscriptionbased and there's pay as you go. They're exactly as they sound,
Speaker A
right? subscription base is a monthly same price subscription generally per user per month for a certain service where pay as you go is like hey you consumed x amount of tokens this month that's x amount of dollars and really
Speaker A
more specifically it's kind of it's kind of more like prepaid. So you you know you buy x amount of tokens and then as you consume those right uh when it gets below a certain threshold you you know you can then automatically buy more or
Speaker A
be prompted to buy more that sort of thing. Really just want to call out right here that generally this is going to be best for organizations that have stronger adoption or aren't necessarily right at the very beginning stages of AI
Speaker A
adoption. It's best for broader employee productivity scenarios with consistent usage behaviors. Where here if you're maybe still in the experimentation stage or have kind of minute advanced agents, you don't necessarily need something like this. Um you can just kind of do
Speaker A
pay as you go that it would be cheaper for you in that sense. Copilot Studio.
Speaker A
So that was that's just kind of general licensing principles. Copilot Studio you can have a pay as you go or a prepaid.
Speaker A
There's a couple of differences between these. The pay as you go is the consumption pay pricing where you only pay for the tokens that you used. And really the advantage of this is the lower barrier barrier to entry with
Speaker A
maximum flexibility. You understand that one week you guys might be hitting development hard and the next week you might not be or one month you are and then the next month you are. that gives you the flexibility to to do that as an
Speaker A
organization. Where prepaid you actually generally can get a discount on the tokens, but this is going to be a more like structured long-term planning approach and really is when a really more valuable when AI is embedded into maybe the more core operations with more
Speaker A
steady use. So, you have pay as you go and you have prepaid. Now, with that, how does this actually work? I I absolutely got questions on this in my exam. So, specifically talking about billing policies and these are for pay
Speaker A
as you go things. Billy policies are what actually allow you to set how many tokens you're buying and when you're going to be buying more, whether it's automatically or manually. You can define them to reset every month, every
Speaker A
quarter, or every year. So for example, if you can have it automatically purchase more tokens for you when you get below a certain threshold or every month automatically purchase more in that sense. You can set up all sorts of
Speaker A
different conditions and filtering. Ultimately if you peak behind the curtain copilot studio usage these billing policies are Azure billing policies. So it's using the same infrastructure that has existed for a while for any of the other existing Azure services. This copilot studio
Speaker A
agent token usage is just tied to an Azure billing policy. You could tie it into an existing one or create, you know, your own copilot studio agent billing policy as you see fit. But nonetheless, like that is that is there.
Speaker A
You are now ready. I don't know what the total time is at this point for this study guide. But nonetheless, you you truly are ready to take this exam. If you watch this back once or twice and truly understand all of the concepts
Speaker A
inside of this video, I'm telling you, you will be more than prepared to take this exam. Don't forget, you can grab these slides to continue your study for free using the first link down in the description down below. And it's really
Speaker A
important that you do because again, you really are just going to have about 1 minute a question on your exam. Maybe maybe a little bit more depending on, you know, the type of questions that you get on your exam. So, thank you so much
Speaker A
for sticking to the end of the video. My name is Griffin Lickfeld, the host of the Citizen Developer channel, and I'm excited to connect with you in the next one, and best of luck on your AB900.
Topics:AB-900Microsoft 365 Copilotagent administrationzero trustMicrosoft EntraMicrosoft PurviewDefenderDLPconditional accessCopilot Studio
