Name: Getting Ready for AI in Web View
Uploaded: 2026-10-07T16:30:41.190Z
Duration: 1 h 28 s
Description: Getting Ready for AI in Web View

Transcript for "Getting Ready for AI in Web View": Hi, everyone. Good morning, and welcome. I'm Heather McClain, principal product manager for innovation here at Blackbaud. And with me today, I have Kara Morris, a senior razor's edge consultant on our customer success team. This session is for Raiser's Edge NXT administrators and advancement operations teams, the people responsible for access, data quality, definitions, and day to day controls that empower your users. Wanna be clear as we get started with the next, forty five minutes to an hour of content that, we're... This is not a case for just turning on AI for every fundraiser. We know that, you wanna be mindful about that decision for your organization. We hope that this will be a practical guide for deciding whether and how it makes sense for your organization, what must be in place first, and what data optimization efforts you should focus on. We're gonna cover things like your organization's opt in to use generative AI, the right permissions you need, data data readiness, and, how to run a pilot if that's the right format for you, and ongoing review to make sure that things are working. So, there'll be some slides. We also have some live demonstrations that Cara is gonna be showing, covering admin setup and ready data and practice. And then I'm gonna be previewing some of, the things that are coming up that are still in development around our automation rules and brand and tone controls. So, just a few housekeeping items before we get started. We have loaded up the resources section. So, as you're listening today, make sure that you click on all of those links. We've got some great blog posts. We've got help tutorials specifically designed for administrators, and enabling your fundraisers to use AI. And we also have copies of the slides. So there is a PDF version of the slides, that you'll see today. And I also just wanna mention that, this session is being recorded, and everyone who's registered will receive an email that has a copy of that recording. We're hoping that we'll have some time at the end for a few live q and a questions. We also have, some folks behind the scenes helping to answer questions while Kara and I are on the stage, but we know that we will not be able to get to all of the questions. So we are planning a follow-up blog post in Blackbaud Community, and we'll include the other questions that we're not able to get to today and the ones that we are, so that everyone has the benefit of those answers and understanding the types of questions that your peers are asking. Okay. One more housekeeping item, for those of you that have been to Blackbaud webinars before, you know about our safe harbor statement. So some items, particularly in the final section, are not fully available in product today, are building now or looking ahead. So plans, design, and timing can change. So please base all of your purchasing and decisions, on functionality on what's available now and not necessarily on a road map preview. So I think, all of that, we've got our, our housekeeping covered. So we're gonna go ahead and dive into the content. So first up, you'll notice that there is an admin guide under resources, and, this is a great tutorial to use to follow along with today. We're gonna talk about, rolling out AI for your organization in three phases. And these are intentional phases that you as the administrator can really govern and manage for your organization. So first is build the foundation. An organization administrator needs to decide whether to opt in to generative AI. And then if you do, you need to create a role and add certain permissions, so that your users can actually use chat for Blackbaud AI and other embedded AI, capabilities within Raiser's Edge NXT. And that means that you can really only assign to certain users that you may wanna include in a pilot. So you don't have to enable everybody. You can, enable yourself, of course, and try it out on your own to get an understanding before running with anyone else. But choose a few fundraisers as a representative group, that you can engage with directly and, get feedback, report problems, and, help to define the workflow that you are testing. And we'll talk about some of those workflows here in a moment. And I encourage you to think about sort of what success means for your organization and what would cause you to pause or maybe even stop the pilot. And, we'll talk about... More about that phase here in a few moments. And then third, sort of expand and govern. Apply learning to support onboarding more members of your team. Keep up to date on the new settings and guardrails you can use to control and to improve the experience. We'll talk about how you can find up find out about more of those a little bit later. So, Kara is gonna lead the setup pilot and product guidance over the next, little while. I'm gonna cover things like definitions and governance and and road map, items here at the end. Please do put queue... Questions in the q and a, as I mentioned, and we'll take them at the end. So let's get started. So I wanted to find a few terms that we'll use, in language today. So... Just so that they are clear, and, not confusing as we go through. We'll try to stay away from any jargon. But these are some AI terms that you're gonna hear, over and over and over again. The first is context. So, context is the information available to the model, to AI, when a response is created, and it has three layers. The first is really, the object being discussed. So that might be a record like the constituent record or a prospect record, an opportunity or action, the work that's really in view for the user, and also the prompt or the question being asked. The second is really your organization. So it's your definitions, things like what your fiscal year is and other settings that we're gonna be talking about today. And the third is the context of the user, their role and permissions, whether they are working as a fundraiser. So if you change any of those layers of context, the response, from the model might change. So while that makes it really hyper personal, for the experience for the user, that that can also add some complexity when you're trying to to test and understand what the experience might be for your fundraisers. The second is guardrails. So these are the controls and limits around access, for, the large language model for chat for Blackbaud AI. Some are built directly into the product themselves. Some are built directly into the model themselves. And then administrators manage other layers of guardrails. So, things like I've already mentioned, like that organization opt in and roles and permissions, your definitions, your code tables, things like that, all impact the guardrails on what AI can and cannot do within your system. And you, as the administrator, really govern and control and configure and document the controls that are available to you, just like you would any other feature in the product, like you would your standard operating procedures for using Raiser's Edge overall. And then, of course, there's human review, and that means a person, a real human, is... Remains responsible ultimately for the result and how you act on it. So AI can produce a draft or suggestion or even provide a quick answer, but a person really needs to check that against the record, and potentially edit it, iterate on it. So don't always accept the first draft as the best draft and decide whether to use it or save it. So review should have an owner and a standard. It's not simply just someone glancing at a screen. So in this session, keep these three concepts top of mind. Context affects the response, guardrails limit an access action, and review is the accountable person deciding what happens next. So we often get this question from administrators, because we know that you really want to ensure an optimal experience for your fundraisers. And to do that, you wanna see exactly what they see before they see it. So I know this can be a little frustrating, but you cannot reproduce an... Your user's experience exactly. That is a real testing limitation, and there are two main reasons. So first, generative AI does not return a a fixed report result. Right? That the same request can be worded slightly differently. The same prompts can result in different answers. And that's what makes it beautiful, but it also can make it frustrating if you really want to understand exactly what your fundraisers are going to see. So you can compare the facts, omissions, and usefulness, but you can't compare identical wording as a test. The second is that context that we just talked about differs. So, the record that may be in view, the permissions that the user has interacting with that record, the organization's data and settings, and the signed in user's role could all affect the response. So that also complicates, especially if you are an administrator, and you have access to everything in Raiser's Edge NXT, and you only want to experience what your fundraisers might see with more limited access. So there's no view this as the user option today in chat for Blackbaud AI. So there's not an impersonation of someone else using that tool. So back on September 22, hopefully, many of you joined us for our InsideTrack on, on generative AI for fundraisers. In that, we really focused on the fundraiser role, and, talked about routine questions, routine workflows. That fundraisers are often, looking to tools like chat for Blackbaud AI to use. So we talked about things like who should I contact, how much should I ask for, Which relationships may need attention and are going going cold? What could I say next? How do I record what happened? So we cover these four different moments. And so, if you were not able to join us on September 22, that recording is available as well. The link is under resources. It would be great viewing for you to go back after the session today and review and hopefully share with your fundraisers as well, so that they can understand how to really incorporate generative AI into their day to day practice. And I do call it a practice because just like any new tool, you have to build the habits of using it, within the workflow, and you have to understand how to, craft your prompts. And we went over in that session some of the tools like the prompt library that you can also find under resources that have ready made prompts, for fundraisers like the ones you see here on the screen, that they can incorporate into their workflow. And so these are a really, really great place to start. So now we're gonna get into sort of what you have to do, to get your fundraiser set up for success. And as we talked about, the first one is, the global opt in. So you need an organizational administrator. So, you may be a solution admin for business at Genxt. You may not be an organization admin. So you may need to engage someone else in your organization to help you with this step. That's because the global opt in, as we call it, governs all Blackbaud solutions, not just Raiser's Edge NXT. So it's sort of the kill switch, for, generative AI features across all of our products. I do wanna mention that, for most razor's edge NXT organizations, we introduced this global opt in after we had already released some of our AI features in the product. So we didn't wanna have any disruption, to your use of those solutions. So if you had ever, in... Interacted with a generative AI feature like AI acknowledgments or chat for Blackbaud AI, your, opt in is implicit. So, we did not opt you out, but you do have the option to opt yourself out at any time. Gate two is prospect insights. So while prospect insights is not required to use chat for Blackbaud AI and other embedded AI tools for fundraisers, it absolutely provides a much richer, deeper experience. It is incorporating our analytics, to help guide those next best actions and recommendations that chat for Blackbaud AI is providing to fundraisers. And so enabling prospect insights for them, will make it a lot more of a useful tool. We realized prospect insights is not available in all regions. It's only available in North America today. So, again, chat for Blackbaud AI, available globally. Prospect insights, available in North America and can enhance the experience. And that's gonna surface things like ready prospects for portfolio review. And so in a pilot, you wanna make sure that, your fundraisers have that access as well. And then we're gonna go over, again, the permissions, and there's several new permissions that have been releasing over the past several weeks. And so we're gonna go over what an artificial intelligence role is and then, how you would, permissions that you need to give to each of the fundraisers that you want to give access to in your pilot. So with that, I think we're going to, get started with the first demo. Thanks, Kara. Thanks, Heather. I'm gonna switch now to the external viewer. I'm going to confirm the organization setting and complete Global AI opt in. I'm signed in as an organization administrator. If you're not an org admin, you won't see the admin option from the top left menu. This will probably be the shortest demonstration ever because it's the click of a button. And that's it. This is the tenant level opt in Heather just described. Be sure to document the decision and the owner. This setting enables generative AI, but it does not assign Prospect Insights permissions or Blackbaud AI chat access to a user. I'm going to enable Prospect Insights next. Again, I need to be an organization administrator to turn this on. If you aren't an organization administrator, then get started button will be grayed out. Now let's set up the security roles for these two features. Under artificial intelligence roles, I can create a new role or update an existing one. Blackbaud AI features grants permissions to the other AI capabilities you learned about in our last webinar. Some of these are premium capabilities, such as action strategies, but more generally available features will also use this role as they arrive, including key insights on the overview tab of the new tabbed constituent record. The important control question with Blackbaud AI chat is whether that respects existing permissions. Existing permissions still apply, so users with different access can receive different responses. Be sure to test this with representative restricted records before your pilot begins. Now let's take a look at adding a prospect insights role under fundraising roles. We recommend leaving view at the very least. You can do this by selecting research prospects. This permission is required to access prospect insight pages and see prospect insights related information in the work center or on a constituent record. Also, at a minimum, users must have the security group access to view constituent and gift information. Managing unassigned prospects is good for prospect researchers, portfolio leads, or database administrators that will be qualifying and assigning prospect service through Prospect Insights. It doesn't matter if you have Prospect Insights or Prospect Insights Pro. The roles and security needs are the same for both. That completes the three checks we set out to show. AI is opted in, Prospect Insights is enabled, and access is controlled through the roles. So we've covered that the organization settings and user roles need to be set up before you can pilot these these features. But we really want a pilot that produces useful evidence without widening access too quickly. It helps to choose one familiar task rather than giving people an open ended chat window. So the example I have on the screen uses a short prompt chain for donor outreach. A prompt chain is simply a few requests used in order in the same conversation. So the wording might vary, but the steps are clear enough to evaluate. Start on the constituent record and confirm you're looking at the right person. Ask, summarize what's new with Robert Hernandez since last month. The fundraiser checks the response against the record. So the test is whether important recent gifts, status changes, and activity are represented accurately enough to help prepare for contact. Next ask, I need phone call talking points for this donor. Your fundraiser decides what's relevant, removes anything that's inappropriate, and edits for accuracy. So the generated text is preparation material, not a script that must be used. After their call, they can ask create a phone call action for this donor. Again, they review the action type, the date, the names, and the notes before saving. If the draft creates more correction than it saves work time, then you need to record it as a pilot result. So we're defining success and stop criteria before the pilot. For example, for this workflow, success might mean users start on the correct record, they catch errors during review, and they save an accurate action more efficiently. You should pause your pilot if important facts are routinely missed, if users skip review, or if the workflow creates risk or rework. Reviewing evidence weekly with a small group is helpful. What facts were wrong or missing? What needed editing? Did the task actually save time? Only add another workflow if the first one is controlled and useful because the objective is to increase prompt... Is not to increase prompt value. It's to decide if our workflow begins and belongs in this process using AI. So during a pilot, we can expect that wrong answers might happen. A wrong answer can come from the request, so the prompt. It can come from the product, or it can come from your source data. If it is from the source data, there's three areas to dig into. So let me explain. Connected identity means duplicates are resolved and relationships among people, records, and households are correct. Current history means recent gifts, actions, notes, consent, and interactions are recorded. And governed means that teams use statuses, definitions, permissions, and review points consistently. So ask these three questions. Did it identify the right person? Did it have the current history? And did it interpret your definitions correctly? You do not need perfect data to begin a pilot, but you do need to log and fix material gaps. So pause the workflow or the pilot when the same gap affects several steps. So what does AI ready data look like? I think we get this question a lot. So we provided three checks to hopefully make it easier to evaluate a recommendation and to trace a problem back to its source. So the first place is to start with portfolios. Use prospect insights as one input when reviewing who might be ready for attention, and then confirm the assignments against your own process. Again, prospect insights is a recommendation, not a must do. When there's an active ask, create an opportunity so that the owner, the amount, the status, and the progress of the ask are recorded. Do not leave active work only in someone's memory or inbox. I'm sure somebody on this webinar has a mouse pad that says if it's not in razor's edge, it didn't happen. And when it comes to AI insights, this is true. Next, review your code tables. Use distinct labels and definitions, remove unused or duplicate values, and map opportunity statuses to life cycle stages. I'll demo this in a minute. The purpose is not cosmetic cleanup. It's to make sure a status means the same thing wherever it's used. Finally, keep notes current. Voice to text can speed up entry after a call, but the fundraiser must read the transcription, correct names and details, and decide what to save. So let's see this in action. I'm gonna review a portfolio, create an opportunity for the ask, inspect the status mapping, and enter a reviewed action using transcription. Be sure to pay attention to the review points between steps. These are the places where an administrator can set expectations and where a pilot can fail safely. So let's start with the portfolio. I'm gonna open prospect insights. Prospect insights will identify prospects that might be ready for attention. I'm treating these as a suggestion, not a decision, and I'm gonna check whether the right prospects are here, whether they're assigned to the right fundraiser, and whether the portfolio reflects the work we actually intend this person to do. I've decided that Bruce Winters is the prospect we want to move forward. But an assignment alone does not track an ask, so I'm gonna add an opportunity. The opportunity is where we track the ask as active work, who owns it, what we're pursuing, where it stands, and what happens next. If you have PI Pro, you'll see this recommended opportunity information that you can accept. I'm gonna fill out my other opportunity fields. Now let's look at the settings behind the opportunity statuses. I'm going to settings, opportunity status. So the values in the opportunity status column are your table entries. The life cycle stage is how Raiser's Edge categorizes your statuses. It's essentially a translation table that tells the system whether the work is still qualifying, moving toward an ask, awaiting a response, or complete. The wording of your statuses can be specific to your organization, and the life cycle stage gives it shared meaning downstream. Be sure to check the count status as column as well because this will decide whether the opportunity belongs in a pipeline or is closed successfully. Finally, I need to record the current context in the record. I'm gonna use voice to text because it's faster than typing after a call. I met with Bruce. We had a great chat. He's gonna talk with his wife, and I'll follow-up in a few weeks. The transcription is a draft. I'm gonna read the transcription, correct the names or details, remove anything that doesn't belong, and then save it. The speed comes from dictation. The quality comes from review. Now my record has three useful inputs, the right prospect with the right fundraiser, an opportunity that tracks the ask in a consistently mapped life cycle, and a current action note that explained what just happened. This gives us a traceable workflow to test in the pilot. So while we're running our pilot, we want to keep support predictable. We want to give our pilot users a regular place to ask questions and report any failures they might encounter. So you can use the prompt library that Heather mentioned earlier as a starting point, but make sure you're requiring relevant context and review. Have your folks share examples that worked and examples that did not, including any corrections that are needed to be... To make sure they can happen successfully next time. The goal is not just AI usage. The goal is AI improved workflows that have been tested. So when something goes wrong, because it probably will or at least it might, collect the exact request, the record, the user role, the available context, and the expected result. Check your organization settings, check your users' roles and permissions, and check your... Then save the example to support because a reproducible case is easier to investigate than general feedback. I'll hand it back to Heather. Wonderful. Thank you so much, Cara. So we've covered portfolio assignments, life cycle mapping, notes, and the data behind different types of asks. So, this slide talks about some of the controls that we have put into place already that are available now and a few things that we are building, in the future currently in early adopter programs. So, and we're gonna go over and, and do some more, review of these things. So we're now in phase three, which is expand and govern. So, hopefully, all of you administrators have had a chance now. We've been, in market, generally available for several weeks on intelligent tags. Tags have guidance that explains what that signal means both to AI as well as to the fundraiser. And a... Intelligent tag is really an input for judgment, and, can help guide the next best actions and recommendations of the AI tools to help fundraisers. Also around goals and your fundraising strategies. So, goals, giving that context, so that word again, context, to AI of the outcomes that you're trying to achieve as an organization, whether those are revenue outcomes, or specific performance outcomes for your fundraisers, like the number of contact reports or the number of actions that they've completed, the types of opportunities that they've advanced. Those, controls are going to help, to guide the recommendations that AI is giving to fundraisers as well, hopefully guiding them to the next best actions that are gonna help them complete, and achieve those goals faster. So, Kara is gonna do another demo of intelligent tags, here in just a few moments, and then we're gonna return and talk about some of those, future road map items, that, are coming up soon that just provide more ways for you as administrators to guide the AI experience for your fundraisers. Alright. So I think. we're gonna switch over to the external viewer just one more time, for this demo. But again, intelligent tags help to move signal to action. And, this, for example, when a fundraiser asks, who should I follow-up with next? Who should I prioritize in my portfolio? Intelligent tags are the tools that are helping, to govern that experience. So I'll flip it back to you, Kara, for the next demo. Thanks. Let me share my screen and show you how intelligent tags work today. This page contains the the tag settings. I'll start on the standard tags tab. These are the ones that Blockbaud maintains and builds for you. What you control here is visibility. You decide which signals your team sees. Start with the signals your pilot needs and add more if users interpret them consistently. Now let's look at the custom tags. This is a premium feature. They're built from my own constituent queries and are evaluated nightly, so they keep records up to date as they change. The pilot should test whether the signal and guidance lead to appropriate review work. So for example, this custom tag is built off a query of somebody that has $25,000 of lifetime giving or more. I can pull a list of my constituents based off their tags, and here I can see that Donnie has the custom major gift prospect tag. As a part of my pilot, I'll want to look at Donnie's record, make sure that the tag applied matches the data on her record. So we saw a quick demonstration of, I'm gonna go back a slide, Heather. Alright. Of the of the custom tags and intelligent tags. But I wanna chat real quick about the still that we have on the slide right now, which is the goals. So this is something that we're working on building. So it's not available today yet. But the roadmap concept is to add more explicit direction. So the goal describes what the fundraiser is working toward, and the strategy describes how your organization intends to move the relationship forward. So in the preview, those future inputs will help explain why particular work could rise and needs attention. So the tag contributes a signal and guidance, and this is just today. Goals and strategies are expected to contribute future direction and prioritization. So, again, this is not current availability, so the design might change a bit. But the administrative questions are still useful for planning. So is the signal in the tag, is it one you want visible? And does the guidance explain a sensible next step? So when goals and strategies become available, will your organization have clear definitions for the outcomes and approaches that you want reflected? Wonderful. Now I wanna cover a few other areas that, are currently in design. So if you join us at our next product update briefing, we'll have a lot more information about these as well as opportunities to join in discovery or early adopter programs around it. So both of these next two slides describe work that's in development. So just use them to understand, sort of how you might, use these tools, to help guide a x... AI experiences and workflows in the future. So the first is around prospect and opportunity automation rules. So we want to make it a lot easier for you to govern routine transitions and make them consistent across your organization and your fundraising teams. So for example, defining what stage or condition, that may trigger a routine update. So things like what happens when a prospect is qualified, what happens, if a prospect is lapsing, what happens when an opportunity moves from one of the stages like, Cara had, demonstrated earlier with the mapping? What happens at your organization when, an opportunity goes from each one of those stages or from when a prospect moves from one status to the next? How do you want to alert your fundraisers? How do you want them to take action? What do you want them to do, whether that's something that you want them to do on the opportunity record directly, like review, the likelihood, or update a deadline? Do you want to create an action for them to do that? Do you wanna link that to one of those intelligent tags, one of those signals that help in the work center with what needs attention, or, do you wanna notify them, in another way? So, we'll have, just like with intelligent tags, a a large, number of prescriptive or more templatized, workflow actions, that you can enable, and then custom, different types of lifestyle progressions that you can configure, based on queries and lists. So, again, nothing changes as far as, human oversight here. You will have the ability to say, you know, do I want this to simply provide a recommendation to my fundraisers for an action that I want them to take, or do I actually want to automate this so that when something happens, it automatically changes something in Raiser's Edge NXT related to that prospect, that fundraiser, or that opportunity. This may be similar to, something like business rules and Raiser's Edge NXT database view, but specifically purpose built for moves management and for prospect and opportunity, automation, and fundraiser enablement in Raiser's Edge NXT web view. Another area that we're actually in an early adopter program for currently and looking forward to, providing access for all Raiser's Edge NXT customers here in the future, is around your brand and tone of voice. So one of the things that, generative AI really excels at and can help fundraisers is overcoming that blank page fear. Right? I know when I was a fundraiser, one of the things that, made me most nervous, especially as a new one, was that I was going to somehow put the, the relationship or even the reputation of the organization at risk, when I reached out and spoke to a donor. I never wanted to put my foot in my mouth. So, one of the things that we wanna enable you to do is to govern and control your brand, and that doesn't just mean sort of your logo and things like your fonts, but how do you talk about your mission? Right? And also tone of voice. So you may have a particular way, that you want to, speak about your mission, when, fundraisers are drafting things like solicitations or, even conversation starters for a cold phone call to a new prospect. So this would be an opportunity for you to, link particular web pages, like the about us section of your website, or, the How to Help section of your website that likely reflects that, that tone that you want your fundraisers to use. And so then whenever a fundraiser is asking to generate content, it doesn't just have the context of that constituent. Right? It doesn't just have the context of the fundraiser and potentially the opportunity and its purpose, but it's actually generating that content, using that brand and tone of voice. The other great thing about the brand and tone of voice tools is that it actually extracts mission stories of impact from those resources that you either upload or link and, attempts to match, those different stories of impact directly to the donor or the prospect on the areas of interest in your mission that they're most likely, to engage with. It's also looking at things within prospect insights, like your donor category and persona, to help match the right content to the right donor to hopefully deliver that right message at the right time to the right constituents. So, this is something that we're very, very excited about and we think is gonna provide a much, more useful, and, experience for fundraisers. So out of the gate, that first draft that generative AI produces is, need less iteration, less, less work to get it ready to be donor facing content. We're also gonna be leveraging the brand and tone of voice controls, within the, contact reports, and AI, voice to text controls, because one of the things that this brand and tone of voice can do as well is extract key terms, from your materials, so that those are are spelled the right way as someone is voice dictating notes about your mission. Again, saving time, reducing the amount of rework that you have to do on the first draft, that generative AI produces. Okay. Those are just a few things, that we wanted to cover today as far as things to look out for in the what's new of research edge NXT, because these are new controls again that you're gonna be able to leverage, to guide that fundraiser experience. I also wanna go over some of the other additional resources that we have, within, the docs or the resources area here. We put the links to make sure that you're you're registered for these sessions. Hopefully, you guys are engaged in our inside track series, which is our product team's, road map preview. So I have another one coming up just next week on the October 13. It's called from reporting to insight. So we touched on goals a little bit today, but, we're gonna go over more of the road map strategy around goals. And, again, those are not just goals for revenue, but goals for things like, fundraiser performance around opportunities and actions, as well around donor participation and engagement. We're gonna talk about portfolio performance, measurement and tracking as well, and, board reporting and some of the future road map work that we're doing around making sure that you can communicate your mission to all sorts of stakeholders within your organization. Like this series, we did the session InsideTrack on September 22, and then we did this follow-up session specific so we could do a slightly deeper dive for the administrators that enable the fundraisers. Same for InsideTrack. So that session on the October 13, we're gonna really focus actually on, the executive persona at first. So that executive role, that might be a director of development, someone who manages fundraisers or fundraising teams within your organization. So, I highly encourage you to encourage your, executive team, your directors of development, to join that session on the thirteenth. And then we're gonna follow-up, with a deep dive that you won't wanna miss. If you are a report writer, or govern reporting, or analytical, pursuits with generative AI, We're gonna cover, reporting that's available today, and automation at a much deeper level, and give you some tools, I call the five whys, to map your current reporting, to AI ready reporting, for the future. Again, we are gonna be following up in community in just a few days. We will have, the slides and the links that are here in the resources as well in case you missed any of those. We'll also have q and a, and you'll be getting a recording to this session. Last and certainly not least, we worked with our Blackbaud University team, and and piloted this before we rolled it out to all of our customers, what we call the peer exchange. So this is a live instructor led, event, that our Blackbaud University, AI specialists put on regularly. It's an interactive sixty minute peer discussion when we really pull up the prompt library, and we work on, how to use those prompts within real workflows, how to iterate prompts and make them hyper specific for your organization. It's a chance for hands on, learning and AI literacy. It is for anyone in your organization. So, whether they're an admin, who enables fundraisers, or for the fundraisers themselves, highly encourage you, to sign up for one of those sessions, through, Blackbaud University. So, lots and lots of, help and resources here. The last that I will just mention is that we do have a link, as well around an AI policy. We often hear from administrators that, they haven't enabled their fundraisers yet because they don't have a generative AI policy in place for their organization. While that's not a requirement to adopt every feature and expand beyond a pilot, it is something if you don't have one that you would wanna consider. And, Blackbaud and our partners have a ready made template here for you that you can use as as a starting place. So if that's one of the gates that's preventing you from moving forward, please, please, please, use that template. It's best practice for a nonprofit organization, and unlock that gate so you can move on to the next ones and get your teams using these really powerful tools. So with that, I think we're gonna take a few moments to, maybe answer any of the live questions, that we didn't have a chance to get to. I think I might have Kara come back up on, the stage here, and maybe we've got one of our backstage guests that's gonna join us too and lob some questions at us. Okay. I am here to lob. But, first, is there anything that, that that you wanna particularly comment in that's outside of what you've already presented? Otherwise, I'll just go into lobbying. Nope. Go ahead in the lobby, I think. Okay. First off, we have... Because the, you know, the time of the webinar is convenient to people on multiple continents, we have people from The UK who are asking about how these things apply to them if they don't have prospect insights, what features will apply to to The UK or Australia Mhmm. from. what you've Absolutely. So chat for Blackbaud AI is available globally. Intelligent tags, available globally. Goals, and strategies will be available globally. Prospect Insights is only available in North America. So we've got Prospect Insights and Prospect Insights Pro. Prospect Insights Pro is available in US today, Canada coming soon. Prospect Insights, not pro is available, and and for all Research and Technology customers in The United States and Canada. So, you do not need to have Prospect Insights to get started using generative AI or for your fundraisers to get value from it. We've designed these solutions to sort of naturally degrade, in the regions where prospect insights isn't available to still make sure that we're answering accurately, the types of questions that fundraisers might have. So, just because it says who do... Who, should I contact today, fundraisers around the world will be able to ask that question, and get a good smart answer to help them prioritize their portfolio and their outreach. related to that question, when you're looking at doing the pilot, it sounded like from what you were saying earlier that it's a good idea to... How how much of the answer do you need to know in order to ensure that you're conducting the pilot rights? Mhmm. Yeah. It's a good question. I don't think you need to over engineer it. Right? But, I think when... Most applies to asking more analytical Right? And we know that one of the things that you want is for your fundraisers to be able to ask questions about their portfolios or about the data in the system. Right? And so, the tool that we use to help chat for Blackbaud AI and our generative AI tools be able to answer those questions are lists in razors at Genxt. And so, you'll want to make sure that you have really clear definitions for custom code tables. Right? So your user defined code tables that result as filters in a list. Right? Because that's what we're going to use to help answer those analytical questions from fundraisers. So having a few common questions, analytical questions that rely on some of those code tables that you've defined, that you understand, that are common scenarios that your fundraisers might ask about your donors or prospects, your mission. I wouldn't say you need to over engineer it, but have a couple of common ones because those are ones that you can ask and get answers to and understand if AI is interpreting your prompt, your question in the same way that you would for your organization. So I don't think I... Kara might have some more guidance, but it's not something you need to over engineer. But have a couple of softballs, set up and and see how. they look. Right? Those... That's a place where you can, have a better understanding of the answers that your fundraisers are gonna receive because we're relying on lists and the filters on those lists. Yeah. And there was a good question, that was posted a second ago here as well about lists and whether we can use lists within a dashboard or tracking opportunities. So we can definitely use lists to, create lists of opportunities, but they can't be linked to a dashboard. Yeah. Not yet. Not yet. We are, under the safe harbor, exploring, ways that we can visualize a list. So being able to craft a prompt for an analytical style question, get a ready list, visualize that list as a as a pie chart, or as a bar graph, and then being able to pin that to a dashboard for reuse. So that is, an area that you'll hear more about in the reporting, shameless plug on Tuesday, in the inside track session. So, it's a great idea for those that asked it, and, it's something that we're we're looking to go from sort of a natural language chat prompt, into creating visualization and pinning it to a dashboard. Alright. And, another question related to, again, international availability. The brand and tone guidelines, Mhmm. will that apply globally? That will apply globally. Yes. Yep. Brand and voice controls, will be a feature that's available to all Raiser's Edge NXT customers, around the world. And it's also really useful. I I I didn't mention, you know, one of the things generative AI is just really, really good with language. So, the tools that we've had today, if you need something translated, if you, if English is not your first language and you wanna interact with AI tools in another language, it... It's very, very good with words and and and language. So, you can use those brand and voice controls in the same way. Excellent. Excellent. And then, there are a couple of questions about the policy template. Did we include. that in the box list? You said... should have been. It was there in the first session. If we missed it, I am so sorry, but we'll make sure that it's included with the blog post that we'll be doing in a few days for that AI policy template. So apologies if it got missed. We had a lot of things in the resources. So, if we missed that one, apologies, but we'll get it posted in community, as a follow-up to this session so you can download it and and use it as a baseline for one for your organization if you don't already have one in place. Alright. But, again, And don't think that a policy template should stop you from a pilot. Right? A pilot is controlled for a few fundraisers, to get feedback, to figure out how you wanna document and govern it where you might have, some data gaps that you wanna address. Generally, having that policy in place before full rollout across your organization, it is a good idea. But don't let it stop you from, experimentation, within a pilot context. That was going to be my next question. So you perfectly anticipated that. So, be developing the template... The the policy, but but continue. the rollout. Yeah. Because that's gonna take a little while. Like, you... Your your senior leadership is gonna wanna review that. Your legal counsel probably wants to review that. Your board will wanna see that like any other major policy for your organization. But, again, having a a a governed pilot... And that's why we present that because, that way you can move forward. You can be making progress in that area in parallel, while those larger compliance and policy discussions are ongoing. Okay. And, another question had to do with how fine grained the controls are over where you can have AI chat show up. Is it on or off, or does it... Can you make it appear only in certain places? Yeah. So chat for Blackbaud AI, is is on or off. So, again, just to reiterate, chat for Blackbaud AI is just one of the AI features in Razer's Edge NXT. There's other things that we've released, like, the voice to text action notes for our customers with access to premium features, things like optimized opportunities and action strategies and some other areas. Those are all generative AI features. And so when you do the global opt out, you are turning them off for everyone and all Blackbaud products everywhere. Right? Again, it's the kill switch. But then each feature will have its own, permissions around them. So chat for Blackbaud AI and using it has a specific permission that you need to grant to the users that you want to have it. Once you do that, chat for Blackbaud AI is available from the top of the top of Razer's iGen XT. It's the sparkle on the upper right hand corner. So it's always available to turn on. And here's the thing about chat for Blackbaud AI is you don't have to be on the place that you are asking about in the solution. Right? So I don't necessarily have to navigate to Robert Hernandez's record in Raiseryside Gen XT to bring Robert's record into context in chat. So if I wanna be on my home page and I wanna type and ask questions about Robert Hernandez and never go to Robert Hernandez's record, I can do that. So it's a little bit of a hard question of saying control where it shows up. Where it shows up really is not the right question to ask. It's what context does it have. And. when you look in chat for Blackbaud AI, you'll see pinned now working with this, and that is what is in context when you are typing. So one of the things that we often hear is, like, it gave the wrong answers. And then when I look and I check the logs, oh, well, you were asking question about this not realizing you were in context of that. So, and how do you sort of clear the context? You hit the trash button, and that clears the context and sort of starts over. So that was a long answer to a short question, but, it it... It's it's a little bit more nuanced than seeing where where it is and where it isn't. Yeah. And it also brings up one more question. We do have a a number of system admins or or, people who are managing databases here who are seriously looking at doing pilots. And one one of the questions that's come up is, is it possible to see a history of usage so they can get a little bit more sense of control over where. AI is being used and how it's being used? Yeah. It's a question that we often get. Today, you do not have access to the specific usage logs. So what are the actual prompts that each of your fundraisers are typing into chat for Blackbaud AI? That's not available. That's not even available to me. By the time that me or, an engineer is resolving a potential problem, all of that is completely redacted of PII, so we don't even see the exact things that were typed in. A lot of that is obfuscated, purposely, as part of our security controls. What we are... And, again, I'll put myself in safe harbor. Now, considering, for the future, is presenting within product, sort of more, higher level usage. So, how many people are, using chat for Blackbaud AI in your organization, how many prompts on average are they entering a day. So having a better sense of AI usage for your organization, but not not current plans to see the actual logs of your fundraisers and what they're asking. Okay. I I think that takes us to the end of our time here. Thank you. Wonderful job explaining or putting into context how you might get ready for, for using AI in the product. I wanna invite everybody again to the, session you had mentioned earlier, Heather, next week's InsideTrack that'll look at reconsidering reporting and how to approach reporting as technology evolves and we have new opportunities, that are presented by AI and and other technologies for reporting. So that should be a very interesting session, thought provoking. We invite you all to join us next. Wonderful. next week. Thank you all for being here today, and, look forward to seeing you all on Tuesday. Thank you so much. Bye bye.