Collaborate on Amazon Invoice Example for Non-profit Organizations with Ease Using airSlate SignNow

Amazon invoice example for non-profit organizations

great and yeah so my name's actually Lauren Stovall and I'm the head of non-profit programs here at AWS which includes our amazing Tech suit program in partnership with techsoup and we have been working with non-profits for over 15 years which is as long as AWS in the cloud has really been around and that's with organizations of all sizes and all missionaries and what we really focus on is helping non-profits transform how they achieve their missions and so I want to take a step back and talk a little bit for those new up to AWS in the cloud what AWS is so the the kind of talk track is that we're the most comprehensive and broadly adopted Cloud platform that includes 200 fully featured services and data centers around the world globally whichever way you want to look at it but what that really means is that you're accessing compute and storage and other capabilities over the internet in a way that allows you to be hyper efficient in how you're consuming types of technology and infrastructure so that you don't have to build it and maintain it yourself so probably a lot of your organizations have maybe servers in a closet in your office or have built your own data center somewhere and you have to maintain that and plan for use and the hardware and all of that kind of stuff and what AWS does is take that element out of it and you actually then consume using AWS services in most cases as a utility like you would water or gas or anything like that so you're paying for what you use as opposed to paying for it all up front and one thing that I think is always really helpful or interesting is that the way AWS actually came to be is through Amazon.com so amazon.com as you can imagine it is a massive um Global web platform and getting bigger every day and so they needed a massive amount of data centers to operate that for the peak of performance which is the holiday season and so they built this massive infrastructure but they weren't using it except for around the holiday season when orders went of course through the roof and they realized wait a second this could be shared and everyone could get economies of scale and efficiency in price and cost if we shared the infrastructure that we've built for our business and so that's actually how AWS was formed and that's now what it is and growing every day and what's what's possible um I would say you know for smaller organizations generally where they start with AWS is things like storage making sure that your data is not on someone's computer who might leave or it might get lost storing that data somewhere where you can have of course security and access controls and all of that stuff but in a way that's hyper resilient and you won't lose it and it's always safe and accessible and then once it's there in the cloud stored there's all kinds of things you can do with it analytics machine learning all of that kind of great stuff and then the other thing and I'll come back to data in a second is compute power so compute is what allows you to run your website for example and that can be you know a static website or a very large web application that your beneficiaries for example use to engage with you donors that use to donate or find out information about your organization and so compute is really the home of the digital experience as well as the back end of a lot of this data processing which I told you I was going to come back to and so that compute power that you can access again when you need it and turn it off when you don't is what underlies what a lot of what we're going to all of what we're going to talk about today which is artificial intelligence and ultimately gen AI which I know everyone is really really excited about and that's what we're going to focus on today but really at the end of the day with transformation whether you're using AI or analytics or just basic business intelligence which is quicksite is our service for that data is actually a massive asset and a lot of organizations I think look at data as a byproduct sometimes an unruly byproduct and the reality is that data is actually an asset a massive organizational asset and we want to work with non-profits to look at the data they have understand what they have visibility into and what they don't and really how to harness that data to have better insight and better decision making and so what we really focus on is encouraging data literacy in non-profit organizations what data do you have what formats what can it tell you and then looking at ways to create cultures around data making sure that when you're making decisions you're bringing data to the Forefront of that whether it's on the cloud or not really to establish that culture of we're going to do this for this reason but it needs to be supported by data opinions are great we all have them but it's also really important to to look at it something objectively and to measure also objectively on the other side to know if something's working or not to further increase your decision making and that decision making can be anything from a programmatic decision you're making it could be a decision about how to make more efficient efficient processes in your organization it could be using the data to Target donors and see how you can can give them more personalized preferences to create that connection with your organization and at the end of the day you can start small it's really about testing and learning and thinking about how to bring data into your everyday life as you go about doing what is the most important thing in all of this which is advancing your mission and going from outputs to from to outcomes to ultimately impact and data is also a key to that right to be able to understand longitudinally over time what is the impact your organization is having and so we have a ton of programs also to help nonprofits kind of go on this Cloud Journey we understand it can seem overwhelming at times on where to start and so we have a ton of different programs to be able to support that and that's everything of course from our AWS nonprofit credit program that we have with techsoup where you can get credits to underwrite the cost of some of your use of AWS annually so every year you can come back and access credits and it'll generally um cover a lot of your foundational infrastructure like your website for example or some storage we also have an AWS nonprofit partner Network which is great and growing all the time same thing with Marketplace so an AWS nonprofit Marketplace this is relatively new but what it is is basically a shop for lack of a better word to go and acquire software and other platforms but the great thing about when you use Marketplace is that it Aggregates it in one bill with your other AWS spend maybe where your website's running and so you're all doing it in a way where everything is aggregated together and you don't have all of these costs kind of all over the place it really streamlines that in one place and then of course there's all kinds of great tools to explore there the other thing is our AWS imagine Grant so that is our annual grant program is a cash grant program the cycle just ended for this year but it'll be open again next year so you can start thinking about your projects but there are two categories momentum to modernize for those just getting started and go further faster those are really taking it to the next level in cloud and innovating with things like machine learning algorithms for example and then we also have an annual conference so the Imagine nonprofit conference that's in DC well it's moved around but it'll be in DC again this year as it was last year and we hope that those that are around can attend or get the opportunity to travel to visit with us though all of that is available on demand after the conference and is now from our last conference in March of 2023 it'll be a similar time this year and so we'll keep everyone posted but really recommend going and checking that on-demand content out it'll definitely give you an even deeper sense of how nonprofits are leveraging the cloud and then finally the AWS powering purpose in the cloud guide so this is based on extensive research as well as data from our imagine grant program that really showcases the state of cloud adoption in the nonprofit space but what it also does is share you can Benchmark where your organization is in your Cloud adoption Journey you can see again how other organizations are leveraging the cloud and the types of solutions that could be very valuable based on your organization goals and so we have a QR code there you can take a second to scan it and again through that you can get the benchmarking tool and the guide spoiler alert it's very long it's like 60 something pages but it is also very very comprehensive and I think thought provoking to help organizations even begin to start talking about what are we doing with our data what are we doing to leverage digital to reach more audiences or enhance the services that we're providing through our programs and so we would love for you to take a look at that and certainly reach out to us on any of these things if we can help you move further in your Cloud journey and the work that you're doing on behalf of your mission so with that I'm going to pass it over to the star of the show EVO to talk about what's possible with you guessed it Jin AI hey thank you Lauren um oh can everybody hear me I think so yes perfect hey so um appreciate that um so like Lauren said if you don't know much about AWS you probably know that we're a cloud hosting service um hosting companies websites hosting your data hosting your business processes right so but we do way more than that so compute and just your virtual machines is one but today what I really want to do is kind of pick your interests and try to have you think Beyond just using um computers as a virtual machine and think what else is possible Right so let's dive into a little bit here so what do these images all have in common so they look like typical photos of dogs and food and mountains but actually all these images were artificially created or generated and they were created with a new form of artificial intelligence that we call generative Ai and so today what I want to do is talk to you about generative AI about kind of a state of the industry and how Amazon can help you use these new technologies to drive your non-profits forward so my name is Evo Janssen I'm a solution architect here at AWS I'm in the non-profits team and like I said it's going to be fun so today um let me click on the right button here there we go so what we're going to talk about today is what is generative Ai and how can we use this for your non-profit and then once we establish why this is going to be useful for you we're going to talk how we can meet that man for you and I'm going to end this with a little cool demo that you might find interesting as well so generative AI is a new type of artificial intelligence that can create new content new ideas like conversations stories images videos music and it's it's powered by a large machine learning models that are pre-trained on enormous amounts of data like internet scale data with a thing that we call foundational model Foundation models and we've seen a revolution in this field over the past couple years especially in the past six months I'm sure you've heard of chat GPT it is everywhere right I'm sure you might also have heard of tools like Med Journey stable diffusion these are some of these generative AI solutions that um Can generate images from just text prompts um so this is not really that new though check GPT stable diffusion they are now new public models and really have brought this to the public but to be honest AWS has been doing generative AI for years for instance this type of machine learning this type of AI that's really what powers the search results on amazon.com and it also Powers Alexa to create a human-like conversational experience right so over time these models become better and better to to fine-tune into your needs and especially in the terms of Alexa to really become this conversational like interface with you that goes Way Beyond the chatbots that you had maybe 10 years ago so I'm going to do a little bit technical here but I'm going to bring it back into why this is useful for machine for for non-profits um machine learning has been around for a long time and with traditional machine learning models you need to do a lot of work to actually make it work for you you would for instance have to train your data with like label data so for instance if you want to build some artificial intelligence that can recognize houses you have to feed it hundreds of thousands of images of houses and after that it starts recognizing other houses or what you can do is you can give it a whole bunch of existing shopping data like we do on amazon.com like hey people who bought like you know uh um batteries also about flashlights right and over time these models learn to correlate those two and I kind of recommend purchases for um for shopping carts now with Foundation models this new type of AI instead of gathering label data for each model which is very labor intensive we're using unlabeled data basically what we do is we throw the internet at these models um and basically let them learn on their own right this is kind of like that neural artificial intelligence so this is what Chad GPT and stable diffusion do so these are these giant public models that allow you to to basically lean on the knowledge of the internet and then generate net new things net new text not new code that new recipes or images right so the cool thing about these Foundation models though is that we can adapt them for specific use cases and this is where you have to start thinking how can this apply to my company right we can adapt these models so they can be customized to perform very specific functions that are unique to your business and that you can actually do with very small amounts of data so this is kind of that that shift that has happened over the past 12 months to make machine learning and artificial intelligence really um reachable to smaller companies and and the public at Large so one of these General uh use cases for generative AI uh some of these use cases again were already possible with the machine learning of the past 10 years and now they're completely improved with this new type of generative AI so let's think about chat Bots right you've had chat Bots when you chat with your airline or with your bank and they've been kind of clunky and generative AI really is transforming the space with much much better interactive capabilities so it just sounds like you're talking to a real person right you can see it in the evolution if you have an Alexa or maybe one of our competitors devices even at home um it talks back to you like it like it becomes more and more lifelike over time um so those are customer experience enhancements um it can help customers boost um employee productivity with like search content creation summarization Etc and we can even improve your business operations with like document processing productive maintenance quality control that kind of stuff so let's dig a little bit deeper onto how this can help non-profits and here's the tip of the iceberg on how non-profits could leverage generative AI and really the purpose of this slide is is to get you thinking on how you leverage this new technology right so here are just a few examples let's think about fundraising like these language generation models like chat GPT can be used to generate qualification emails thank you letters or even generate new proposals um member interaction you can improve your customer service uh think about again like these chat bots of of the past they are now completely overhauled become much more interactive and much more smart about your business um grant writing um you can use these these text generation models to draft Grant applications project proposals using data and text from your database of brands a search uh companies especially non-profits they often have a huge knowledge base of data a lot of customer data so let's think on how you can connect this new technology to your existing crms and other systems of Records to to search across your knowledge base uh content creation um there's a couple of public AI platforms it can turn blog posts and written content into videos automatically so that's new uh Lumen 5 synthesis that AI these are some of the public companies that help you with this and we have tools that can create um a targeted fundraising emails like I said volunteer management um so we can use this this predictive aspect of of AI to help you identify and engage volunteers match them opportunities track their purpose Etc and one that I think is pretty cool is as Outreach um uh um uh sorry multimedia I think it's actually happening right now whereas closed captioning that you can turn on the zoom call right now so that's the type of artificial intelligence where it picks up your voice and helps summarize video or transcribe video to make them more accessible so really what I want you to do is to think about all these things that are possible right so here at AWS we're all about what we call the art of the possible and we'd love to hear from you what kind of use cases you can come up for this new technology just a real quick example here um like I said a transcript summarization or article summarization for for grant writing this is some stuff that our partners call here on AI 21 do um worktune is a cool tool that actually helps you improve your writing by making it more engaging um image generation this is kind of what stable diffusion does this is actually in the tempo by stable diffusion 2.0 um this living room that you see doesn't actually exist it's completely computer generated right you can think also about taking two different concepts like a human face and a VR glasses and then actually ask the model ask this Ai and you combine those two and actually give me a human face with a VR goggle and showing off AI can do that so we're not new to this um we have here at AWS have been doing machine learning and have been taking this to our customers and making it easier for our customers to do for the better part of a decade now we have over a hundred thousand customers using AWS for machine learning alone including many many non-profits cool so I think at this point it's clear that this is going to be useful this is going to be transformative for the industry for all Industries including a number for industry right and we've seen these public models check gbt stable diffusion that are out there that anybody can use you can download an app um you can log into a Discord server and start creating these images um but think of this right all that data is publicly available and that's challenges on how to apply that to your specific um specific company so there's a lot of privacy concerns and I know there's questions in the chat so I'm going to address those right now there's privacy concerns about feeding your data to these public models like charge EBT right because think about it these public services will learn from your IP that you're typing into it and actually might give it back to other customers and you really want to avoid using these public models from for your targeted Enterprise um Enterprise applications second you also need to customize these models with your company data what good is a chat GPT if it if you have it on your website but it can't actually answer the FAQ for your specific company right and all this has to be done in a secure way and in a cost-effective manner so let's see how AWS specifically can help you take this concept of genotif AI and make it make it applicable to your company so um first of all how do you even get started right I mean I know that some of you are new to AWS that's probably some of you that have only seen chat DBT uh in the news so how do you get started with that right so we actually made this available to you so what we have done is we have curated a selection of these Foundation models as they are called from AI 21 from anthropic from stability Ai and even some of our own models and we make those available to you through our AWS services and we'll dive a little bit deeper on how we do that right second you need to customize it right so it's one thing that you have access to these these models in a in a secure and private way but then you need to be easy to take this base Financial Foundation model and then build these different differentiated apps on top of it that use your own data and again your data is your IP so this is a secure protected and private during the whole process you need to be able to to be in control of your data so you know it remains private and confidential right so and we do that so if you know a little bit of AWS each little what we call virtual private Cloud within AWS it's completely private to you you might be running on a on a wreck and some data server right next to what your bank or next to Netflix but we have segregated this um the government uses it as well so you can trust it's secure um any data that you store in your virtual private Cloud will remain there and will not be shared across any other customers and that's really important cost effective um basically everything we do at AWS is space you go and you use something uh in AWS and when you're done with it you can turn it off and your cost goes all the way down to zero right so you don't have to pay or get locked into upfront contracts you go like hey I need a virtual machine for the next month or I need this gen AI model for the next five minutes and you can build per second basically so that's how we do pay as you go we even developed our own chips um our own graphical chips um to make it even more cost effective to you okay so you need to be able to use this quickly right so like I said in the first bullet point and flexibility we make these things available to you and how do you even get started to using those so we have a number of different Services they're called Amazon sagemaker I'm going to talk a little bit about Amazon battle Rock which is a new service for Gen AI um these Services allow you with very minimal coding sometimes even no coding to implement these generative AI Solutions into your websites and into your business processes talked about no coding Solutions so what we do here at Amazon as we start building taking these these these bits and building blocks and bits and pieces and we make them available to you in a completely managed way we already do this with OCR document scanning with text to speech speech to text now we have a new service completely built on General diff AI gold code Whisperer so if you're a development shop building code I'll show you a little bit demo of this a little bit later it can actually write code for you this is crazy how easy it is okay so um for those of you that do know about AWS and services this is the net new that we've announced for this year within the context of genotype AI we have two new Services well really one new service one was there already a little bit so uh Amazon Bedrock is a completely new service that makes it super easy to you to build and scale new generative AI Solutions with these Foundation models from our partners um and as our Amazon sagemaker jumpstart um which allows you to basically take this um these models and modify them for your specific uh use case I'm going to show you a little demo of sagemaker jump start in just a little bit um let's see a special specialized uh chips I'm going to jump over them it's really cool we have cool technology and Amazon code rest for is this new managed service that allows you to help build your applications and your mobile apps faster okay let's dive a little bit into each of these before I dive into my demo um sagemaker jumpstart it's um a tool and you'll see it in um a live in action in just a few minutes uh you get full control over your infrastructure you get full control of which Foundation model you want to use how you want to customize it so it's it's quite easy to use um we have a long list of publicly available Foundation models that you can use and run and customize and then eventually also deploy into production into your into your ecosystem lots of different models we have proprietary models publicly available models the proprietary models are they cost a little bit more but they are super accurate um and that's why that's why they charge a little bit for them but even the publicly available model they still offer complete visibility and control over all the parameters and you can still customize us as well so sagemaker jumpstart we'll see in a second it's a little bit if you're not used to coding this might be a little bit of a lift for those of you that are a little bit familiar with with writing code to see how easy it is to generate um to use these these Foundation models then we have a new service in preview called Bedrock which is going to make it even easier to use General Cliff AI it's completely serverless so no worrying about CPUs or virtual machines or CPU hours and and costs like that you really just pay uh per API call um so with betrox serverless experience you can easily find that right model easily customize it again privately and securely with your own data and then integrate and deploy it into your applications that you also have running on AWS like on a web server or on a database or something like that like I said we have partnered with top AI startups to bring these Foundation models to you in an easy way from AI 21 and Tropic stability AI um and even ourselves so even Amazon we've built Our Own Foundation model um that basically based on 20 plus years of our own experience with for instance amazon.com search or or Alexa and then Amazon code risk for this is just a really cool application for um for completely managed uh a general fi there's no customization needed so what we've basically done is the AI coding companion it helps you quickly write secure code by generating full function code suggestions in your favorite IDE in real time based on this on commands right so I just type in my code like hey I need to parse the CSV string and throw the list but the normal is that you type enter and hop Amazon codeword actually outputs all this code right for you this is really powerful stuff you might have actually seen this if you're into coding that this is something that we do uh that chat GPT can do as well but again think about it right chat GPT that is a public service and got its training from the internet and as we know the Internet isn't always right so the results you get back are not always right either so we did with code risk or strain it on specifically curated code to make sure that we get really high quality code back cool so I talked a lot so what I talked about is basically um was genotype AI right we the code generation we had text to text we had text to image and we showed how you can apply this to a your star a non-profit in a in a secure and cost-effective way so I want to kind of give you a little demo of of sagemaker jumpstart and and show you how easy it is um to do that so I'm going to switch to the console here so if you've not logged into Amazon then just go along for a ride and enjoy if you have logged into AWS before um then this might look familiar for you right so um in the console so what I'm going to do here is go to Amazon sagemaker so you can see a lot of different Services here ec2 those are virtual machines but most people think of and I think cloud computing um uh there's other things storage S3 but I'm gonna go to Amazon sagemaker here and so what you see here on the left is also jumpstart at the foundation also click on that so right here in our console you see all the foundation models that we have partnered with and that are available to you inside um Amazon sagemaker there's hugging face text generator there are um some GPT stuff as well um and all the way at the bottom here is a stable diffusion so stable diffusion is that um text to image generating a tool a computer model here shows you the USA and so really what this does is once you click on open notebook in studio this takes a few minutes so what I've done is instead of clicking on here I'm going to switch to this new tab while this is already opened and so what this gives you is a playground to play with this model to kind of try it out to customize it even so what this whole is basically tutorial it's a lab right so it goes through all the steps you need to do to um to get this model up and running again in your own virtual private Cloud so nobody else has access to this particular one this is all yours and then we'll show you how you can customize this um if you're new to coding this is a lot of python code um this is why we came up with Amazon Bedrock which actually makes this a lot easier and Amazon better work I don't have a demo for that now but it's in preview right now um so basically it's it's really a tutorial so you can follow along um and really all the coding has been done for you you just have to click on yeah let's do this let's do this um and really what it does all the way to the end is build um that that that image generation interface all the way in your own um in your own account so this is completely done in my own sandbox again nothing um anywhere else is not public it's just for me but once I'm done with this I can type problems like draw me a college an impressionist style and sure enough the computer now builds this because the computer being trained the model being trained on this internet scale data knows what a cottage is right and knows what impressionist style is so this is a net New Image completely generated by um by computer so so far this is possible out on the public info net as well right so you can actually go to um to stable diffusion's website sign up for a free account log in and start generating these images yourself but like I said anytime you generate an image that gets that's a public image of what people can see it too and now all the prompts you give it uh the model kind of learns from that as well so that is not apply for data so what I want to show you now is how you can take this stuff and customize it with just your data as an example what I did here is I took a couple pictures of a of a dog dog is called Doppler it's very cute but this is this is a dog that that the model doesn't know yet about right these are this is this is not these are not public images of the dog that the model knows about this is a completely separate dog his name is Doppler and basically what I have is about let's say you know 20 or so pictures of Doppler so very cute it's not a question is how can we take this giantly trained model that took like a year to make and how can we make it so it now knows about Doppler so it can start actually giving us net New Image generation you just offer right so I have these pictures a little bit of a a little uh some metadata here that says hey this dog is called Doppler and yes it isn't actually a dog and so when I have all these files in my directory what I do is I upload them um scroll down a little bit here so I know is I basically upload them to Amazon so I upload them to S3 S3 is our um uh storage within AWS so everything that we that you store our websites data Etc all goes into S3 so I put that in there and again you can upload your own images as well right um but I chose these pictures of Doppler right here and then what you do is you basically train the model on a um on this differential data and we call a trans folder on it so again all the code is right here for you you just execute it if you want to try this let's talk about cost for just a second right how much does this cost so you only pay for what you use so the model is already pre-trained we actually make that available to you at no charge if you want to do this additional incremental learning you'd have to pay for the um the the time that you use these these CPUs trainers 20 pictures uh takes about maybe like an hour or so for this for for this to learn about this dog about one and a half hour I think I I did yesterday to train this model um the CPUs that we use are about let's say three dollars an hour so let's stay say it costs about five dollars in total to um to teach this model on how to think about this dog So Randy last night and now it's now it's ready to go so now I can actually give it a new prompt and go like hey I want a photo of Doppler but now with a hat if you think about it right nobody um uh none of the pictures I had actually the dog had a hat so what if I just have the model do this takes about 10 seconds to generate there you go so this is a completely newly generated picture um of that dog it does look like him right and now it has a head so this is a completely new picture and the model is trained on my private data which is completely different from um using these public stable diffusion services on the web I could do something else this is this is a red hat can I do a let's see kind of do a yellow hat let's try something else life always dangerous if you do demo live there you go now it's a yellow hat right maybe I want him on the beach it's about 10 seconds and so the system the cost of this by the way so again think about it right this takes about 10 seconds to generate like I said the CPU that Amazon that areas charges you to to do this is about three dollars an hour and so think about how much it costs to do 10 seconds of image generation there we go now we have Doppler our our trusty a dog with a yellow hat on the beach if you want more example I actually want a queso uh style painting instead of a picture let's see if you can do that so again each each generation here is Little League pennies right there you go Isn't that cool so um the the model stable diffusion already knew about Picasso already knew about the style of Picasso but we just taught it uh um with this transfer learning about my dog Doppler and now it can actually make the Gasol style paintings of Doppler on the beach with a yellow hat okay so let's let's go back to our slides and let's see what we've learned here so what have you learned here right so this is stage make a jump start so we used a pre-trained foundation model in this scale case stable diffusion which already knew a lot about the world and knew what a beach looked like not what a dog looked like and no other a hat looked like but it didn't know Doppler um with sagemaker you code in Python and you pay for CPU use by the second so again took about like maybe like five dollars to teach it about my dog and then it takes literally pennies for 10 seconds of image generation Amazon Bedrock is a new service as in preview right now so if you're interested in that we'll have some QR codes on the screen in just a second to teach you about um about new uh Services it will be serverless so it'll be a lot less python building um and the way that you pay for it is basically by API call rather than by um uh um having to think about CPUs and stuff like that and it's customizable with your own data right and that's so important for for you as a company how can I do this um targeted for my use cases while I retain control of my data this is a graphical example because I thought it would make a nice demo with a cute dog right um but think about the use cases that you can also do right earlier we talked about um we talked about grant writing right so if you have a database of of um of proposals you can feed that into these text models these chat GPT light models and actually start building net new content trained on your data um but then for your purpose only based on that data you can train it on your customer database you can train it on your um on your FAQs Etc and all of this is space you got pricing I think I mentioned that right so um right after this demo I'm going to shut everything down and my cost from that point on will be literally zero so this whole demo basically uh the full price of this was maybe like a few dollars you can try it yourself if you're used to Amazon uh go to the console go to sagemaker jumpstart and pick the stable diffusion 2.1 model um it will actually um uh like populate your your editor with all that code you don't have to know python you can just click through it and you can see all kinds of examples of auto-generated images and you can upload your own images as well and see um how you can customize these models for your particular use case cool so um that kind of concludes the demo I'm gonna have some some links here um these slides will also be available on a taxi website starting I believe tomorrow so if you don't have a chance to snap all the QR codes right now the slides will be on um on techsoup and on YouTube pretty soon for you to to uh to see them again and so with that I'd like to say uh thank you um well my pleasure to demo this to you I hope you had a it was informative um and let's start thinking on how you can use this for your use case thanks Evo and we do actually have some questions if you don't mind not at all all right the first one is from Tom and his question is regarding stable diffusion and image creating with AI how are copyright and the artist rights being considered in the images we create is it safe for our organizations to use images created with generative AI that's a good question um so in the do I still have that up I think here in the foundation model if you click on view model I've already had it somewhere here right so the utilize right there um I'll be very honest I have not read through the whole you know myself but we do provide with each model exactly what kind of um uh you know requirements there are of using the generated images from here so to be honest I have not read threat through the whole Ula so I don't have a specific answer to your question but what they're going to say is you can probably find these answers in the Euler that we write with each model okay and and then Brittany had a question to add to Toms which is we have been using AI to generate news articles and other marketing materials we have not been citing the AIS we use should we be giving credit to the AI program we use yeah I think that's going to be the same answer I I'm not a lawyer I'm a technical guy who knows how to make a cute pictures of dogs um but I would say let's read through the EULA on exactly what the copyright implications are and I I to be honest I'll go read through that later myself and maybe I'll learn something for the next time okay this one's from anonymous regarding the use of large language models or llms how protected is our personally identifiable information for our participants yeah so that's exactly what you've what we've shown you today right so basically what you're doing is you're spinning up a copy of that model into your own account so think about it this way your data is already safe in your own VPC your virtual private cloud and you're bringing the model to the data whereas on the public you take your data to the model on the internet so really by having your data already private and then running that model the llm inside their own accounts none of the data that you have will be sent back or used to to sort of train the original base model it all stays local through this concept of transfer learning and then someone asked and if you have something Evo great all I can answer it to start it says can you share some reference accounts that non-profits of nonprofits that are using generative AI today the short answer is that generative AI is relatively new on the scene we have a number of organizations that are not yet in terms of that use case and story uh referenceable we always of course with any of the organizations nonprofits we work with ask their permission when we share any reference so for the moment not referenceable however if you saw Evo's slide earlier you can go back to it with the NASCAR Slide with a lot of organizations on it those are all non-profits that are using AI or and or machine learning in some way today and so that's organizations as large as The Nature Conservancy new organizations that are just growing like stop Soldier suicide and across you'll see all different Mission areas on be the match National Maradona program is of course in the health space stop Soldier suicide mental health and suicide prevention major Conservancy all of those and so as you see hard American Heart Association there's a number of organizations again and all over in terms of size it's really depends on the use case that you have for AI and ml it doesn't solve for everything but if with the right use case it can be a very powerful tool and so we certainly have on our website which is our AWS nonprofits website a number of case studies and use cases around AI we have a non-profit blog that does the same and certainly you can reach out to us directly via our website and we can go into much more detail about the various uses of AI and case studies that exist as well as the use cases and deeper in yours in terms of generative AI um Evo did you want to add anything to that before I move on no I think that was exactly right thank you perfect okay this one's from Tom again thank you for all the questions Tom what metrics do you have published on the accuracy of code Whisperer and your other models let me pull this slide up here I didn't show it in the terms of of speed um so uh does accuracy uh I'm not exactly sure but we did have some metrics around that um so we did a preview we did some challenges and some some examples because a lot of this also relies on how the actual users used code whisper by actually typing uh um useful and correct uh prompts to actually generate that code what we found is that participants who used code quest for 27 more likely to complete tasks successfully I know that's not 100 question I don't have a specific stat on actual code accuracy and I'll see if I can find that for you all right and last one does AWS AWS integrate with other systems such as QuickBooks Online is there a donor management component are there other demos or support if we are interested in joining AWS right so um to answer that there we have a large ecosystem of Partners and we basically partner with pretty much every company out there um and we build a lot of Integrations and customers build Integrations with AWS right so AWS is more than just stage maker it is the virtual machines it's databases it's all it's 200 plus services and all these Services we are what we do is really interconnect all these services with your existing systems that might still be not in the cloud they might still be in your own data summer um so we have a huge expertise in building these Integrations yeah and so yeah right yeah go ahead and to add to that point there's a service called app flow which helps build those connectors from your solutions to the cloud if they're not on the cloud today but for example a lot of those connectors already exist the Salesforce connector already exists for all of their capabilities the blackbaud has a number of connectors for example if that's a tool that you use and so so the connector some of them are pre-built some of them need to be built depending on the situation but all of that is possible integration is absolutely possible and it's frankly recommended because that's how you get your data all in one place and then you can really take advantage of capabilities Beyond even the platforms and software you're using whether it's analytics and other types of capabilities quicksite to do visualization and really create that one source of Truth for your organization where you can tap that data with the right permissions as you need to to look at it for different reasons the CFO might have a different reason and question of that data than your fundraising team and so that really really bringing all that data in one place is a highly effective way to create insight for your organization to make data-driven decisions yeah I'll at one point to that Lauren so my role is a solution architect so apart from doing webinars every now and then really my role is to talk to our customers to make them help them integrate all these Solutions and make them have them make sense of all our Solutions so I'm not sure who you're with but please reach out to us and we'll be more than happy to set up a conversation with the solution architect that um that covers your your specific company or domain yeah perfect thank you Evo and then for those of you I know a number of people and there was also kind of a question in here about training and capability and so I will say that we have a ton of free digital training on a variety of AI of course but everything from foundational what is the cloud all the way up to really Advanced use cases so absolutely recommend checking that out and again the partner network is super super important we absolutely understand that nonprofits are Change the World Companies and they're not always deep technologists and that's okay that's why we have cultivated a really great network of organizations that are dedicated to supporting non-profits and helping build that capacity and that's what we help with as well to Evo's Point our solution Architects are here to create that guidance and in talking with the solution Architects based on what you're trying to do they're really a great resource to help you figure out that Journey what training should you take what partners are good to engage with so you don't have to do it out there all by yourself it's really we serve as a trusted advisor in that capacity so you can get to the right place quicker this was so excellent uh super super informative I mean wow I'm looking forward to the next one and I'm not just saying that this was a lot I learned a lot um very different from your previous webinars um just they just keep building and building and building so if you if you missed a previous AWS webinar go to our YouTube channel on techsoup YouTube channel and I know some people here may not have been um technically a Savvy but this is the direction that the world is going to so I know this webinar is going to be super helpful for you if not today in the future and we look forward to having the AWS come back again Tom I mean Lauren uh lock in any last words Evo thank you yeah thank you and we are happy to talk more I to reach out I know that has been a question go to our AWS nonprofits website there's a start a conversation link that will come right to us on our team who will understand what you're looking to do and get you to the right people who can provide that guidance mm-hmm awesome have a great day everybody bye-bye foreign