008 Using AI to Rewire a Whole Nonprofit | Hope Rises International President and CEO Bill Simmons

In one month, CEO Bill Simmons shipped more than 15 working applications and 70,000 lines of code to his team, without adding a single software license or outside developer. 

He did it with about $25,000 in internal time, replacing software that would have cost over $750,000 to build. Most nonprofit CEOs are still asking whether to pay attention to AI. He's already restructuring how his organization functions by using it.

Bill Simmons is President and CEO of Hope Rises International, a Christian global health organization working in more than 50 partnerships across Africa and Asia to address neglected tropical diseases. Annual revenue: $11.3 million. 

This is the first time I've seen how AI can transform the processes and outcomes of the nonprofit sector.

Transcript

Introduction to Hope Rises International

Adam Jeske 0:05

Nonprofit CEOs are constantly making consequential decisions, often under pressure or in isolation. I'm Adam Jestie, the Nonprofit CEO Advisor, and this is the Nonprofit CEO Podcast, where CEOs talk about the decisions they carry. Today, our guest is Bill Simmons, president and CEO of Hope Rises International. Bill, welcome to the show. Hey, Adam, it's great to be here. Looking forward to our conversation. Tell us a bit about Hope Rises International.

Bill Simmons 0:33

Sure. So Hope Rises International about a year ago actually went through a brand change. We were American Leprosy Missions and we started in 1906, incorporated in New York City in 1917. So we've been around for a few minutes. But we continue to work in a category of issues that affect people and lead to stigma from diseases. And so I always say it's the sort of things that people in a community would call leprosy, whether or not they uh know the underlying medical diagnosis or not. And so there are several diseases that lead to that sort of stigma and separation, isolation, and disability. And so we work with those diseases around the world. And hope rises better captures for us what we're trying to do in the world. We're trying to reach people affected by these diseases and not just leprosy and see them discover hope in their lives. And so that's what we do around the world.

Adam Jeske 1:35

And you've been leading there for a while.

Bill Simmons 1:38

Yeah, I have. Well, actually 16 in December this year. So yeah, it's been tell people I came from the private sector. I was the CEO of a couple of different uh retail chains uh in my career back when uh bookstores outside of Barnes Noble and Amazon could uh could make a living. And uh came to this sector. I was a missionary kid and uh always had some desire to do something in the world that uh was different and not just for profit. And so came here and I tell people it's it's my dream job. I just didn't know I should uh should have been dreaming about this one specifically. Uh but I've loved all all 15 plus years that have been here. Yeah, it's been great.

Adam Jeske 2:21

So which consequential decision would you like to start with that that you've needed to make while leading at Hope Rises?

Bill Simmons 2:29

Yeah,

The Significance of Rebranding

Bill Simmons 2:30

I you know, the the brand change was certainly and remains a significant decision. Um and I think it's it's a lot of it is in the mind. I I would I'm probably not I'm not gonna land the plane here, but I can't dismiss it. I need to at least acknowledge that a brand change is really significant. Um we didn't change fundamentally what we do in the world. And so it wasn't a shift in product, as it were, it was a shift in trying to communicate what we're doing uh to today's world. And the the words American leprosy emissions all are a little stale in the current climate, and we had to accept that at some point. Uh, there are other organizations in Europe, for example, that also were leprosy missions of some sort historically, and uh I think out of the 13 that I'm aware of, uh 12 of the 13 have all removed leprosy from their names and have moved on from that. And so uh we felt like we needed to do that too. So it's so it was sort of terrifying, of course, to think about removing um some of the key attributes to who we are. But uh, you know, I think that's it's turned out not to be the most significant challenge we face. There are other challenges, I think, that are in our landscape. And so uh it was it felt like a big decision. I'm not sure I would put it as the most significant decision that I've had to make as a as a leader. Uh, but a brand change is a big one for sure.

Adam Jeske 3:59

How long were you considering that or wanting to do that before you actually entered kind of formally into the process? Yeah.

Bill Simmons 4:07

I mean, we we we were well into the process when COVID started. And so, in fact, we had a deck sort of a process ready to go and to pull the trigger on, and then COVID started, and so we decided that that was another disruption we shouldn't enter into the mix. And so we put a pause on it. And uh so in some way, I you know, I guess we we were in the process of thinking about it for at least five years intentionally uh as we launched.

Adam Jeske 4:35

And as you approached the decision about uh within the brand process, were there competing approaches or strategies to how you might represent the organization?

Bill Simmons 4:51

No, I you know, I think the biggest challenge we face is the crowded marketplace for names, actually. Um it was can can we both find a name that is available and describes who we are and what we're trying to do in the world. I think that's the growing challenge that an organization changes, uh faces in a brand change. So we knew we needed to change the name and so it was and we didn't have a lot of like argument or discussion about um about that. Well, I shouldn't say that we we had a lot of discussion, just not

Navigating the Decision-Making Process

Bill Simmons 5:19

a lot of uh disagreement in the end. We came to the same place. Um but I admittedly I would say it's it's it's not the perfect name because those don't exist anymore.

Adapting to Technological Changes

Bill Simmons 5:30

I think the most important decision that I have made and am a part of is meeting the landscape change that technology is is putting in front of organizations, how to meet it, meeting it well, um, adapting to it, getting in front of it if possible. Uh and uh that is the most I I am convinced that that I'm on the front edge of that. It's out on the jagged frontier of this technology. But I am so convinced that the way that work is going to be done and how organizations, nonprofits or otherwise, will function is going to shift in the next months and years, already has that the very thing the CEO's there to do, which is to carry the business and the vision of the business forward, if you're not evaluating how this is going to change the landscape, and then being a part of that change, I think that introduces a lot of risk. So the boldest change and decision I'm having to make is actually fully implementing and adopting technology in a way that is transforming our organization

Innovative Solutions through AI

Bill Simmons 6:45

and it's transforming it live uh daily in front of my eyes. And we are doing things that some might say are risky, uh or well, and say they do and carry some risk with them, uh, but we are introducing things into, and I'll say specifically, to organizations that are in the you know, that are less than $50 million range. I think the uh resource availability over $50, 100 million changes. Uh and so you may be able to throw more humans at change. And in our case, in smaller organizations, mid-sized organizations, we tend to have less free resource to easily allocate to adapting to change. So it's big organizations in the past that all of a sudden could take advantage of new changes in the marketplace. And with today's technology, AI-enabled technology, organizations and CEOs that are willing to do some bold things can execute the levels of change that would not be financially feasible or feasible from a time management standpoint, uh, even a year ago uh wouldn't have been possible. And yeah, what what we're doing is pretty cool. It's been quite an interesting ride. And and when I say interesting, I'll I'll finish by saying that what we've built is is less than a month old. And it's pretty I I really pretty I'm pretty amazed that um that we've even done it. And uh and we've just begun, so to speak, uh, with the total transformation of of how our um organization works and functions.

Adam Jeske 8:31

So that's very intriguing. I want to um operate on a couple of levels in the conversation uh and take us back to the beginning. Um so the levels would be you personally, your leadership, the inside your head, and then also organizationally, what are the changes that are afoot? What have you already built? What do you think the frontiers are as you look out across the sector from both within Hope Rises International, but then also elsewhere in the sector? When did you start paying attention intently to how AI was arriving and what it was beginning to permit?

Transforming Organizational Workflows

Bill Simmons 9:12

You know, so I've always been sort of a a tech architect, I would say. You know, I've never been like a coder. That's not my background, even though I I toyed with that early in my career. But I was always uh an IT infrastructure architect. That's how I started my early career. And so I would say I've always had a penchant for uh seeing how the puzzle pieces went together from a tech stack discussion. But I haven't operated in that sphere in a while. There's a great substacker, Jason Crawford, he talks about this recently, because he's a nonprofit CEO now, too, but uh he's one who also sort of grew up in a more IT enterprise uh area, and now all of a sudden he finds himself using Cloud Code and doing all sorts of things he never thought were possible. And I identify with that sort of um of a current uh mind frame. And so what I found myself doing as um AI tools reached the marketplace was you know, I I started using them. And there were two uh pivotal moments in the story for me. One one was um I was going on a trip with my with my kids, and I waited too late to um to get tickets to Harry Potter World and in the UK. And uh but I but I knew that a lot of the tour companies put tickets back on the you know into the pool as they got closer. And so I uh I asked, it was Chat GPT at the time, I said, well, can't we just couldn't we create some sort of software that go out and would scrape their site and would send me a text when tickets are available? And say, well, sure. And so anyway, walk me through the steps. Long story short, and um in about an hour I had a running app that would send me a text and let me know when tickets were available. And the moral of the story is that yes, we got tickets and we got to tour Harry Potter World and in the UK. Um, and so that was really that gave me a window into this core idea, I think, about AI, and that is that if you can think of it in in the realm of digital space, it's possible. And that was when I first realized anything I could start to imagine was fairly, fairly possible. Then I would say I think I ran out of frontier, and this was about um almost a year ago. Um and I started to try to use AI technology to uh completely change um some internal infrastructure. And it turned out that it didn't have from a coding capacity the ability to really do it. And so I sort of sat on it. I ended up using AI tools the way most people do, as sort of, you know, um superpowered assistants in some way. And you know, I tried a lot of different tools and lived in the chat GPT space for a long time. And then in the second shift, you know, what's happening in my mind is your question, is in January I started using Claude, and I don't even remember why. Uh I don't remember who I was talking to and or who I was reading. I finally said, you know, I'm just gonna have to try Claude. I tried it once, but I wouldn't really use it. So I started using Claude, and I really

Building a Cohesive Brand Identity

Bill Simmons 12:22

liked uh I liked it as an organization leader because Claude's output looks like organizational thinking and it looks like organizational documents you're familiar with. So it's more likely to give you a Word document when you ask for something or an Excel document and or a PowerPoint that actually works. It's more likely to give you a product that is a part of a workflow in an organization, number one. Then we had a problem, which is when the turning point happens. So the turning point was our marketing agency made a mistake at your end tax receipts, and they sent a bunch of donors tax receipts that said thank you for your zero dollar gifts. Um so the so the phones were ringing, and our team uh we used Salesforce uh for our our our CRM, and so our team was taking phone calls, trying to copy and paste information into a Word document, and you know, it's this whole laborious task. And so I went to Claude and I said, What do you need for me to solve this problem? And said, I need two things. I provided it those two documents, and in 15 minutes I shipped software to our team that they could open up in a browser, could put in the donor's name, and they could print uh a PDF or print a tax receipt and mail it, either one. In fifty in 15 minutes, I had published software. Man. So that was the first domino was realizing that the tools are out of space to bring real tools and solutions to people on the front lines in organizations. Um is what I did a revolutionary. Absolutely not. If I could afford, you know, four or five developers, then we could have coded that. No, it probably still would have taken a day to do it. You know, so it's not impossible, but it's pretty inaccessible to organizations below 50 million, 100 million dollars. And so I immediately saw the potential of where we were in the landscape. That's when everything changed. That was the moment that led to where where we are today. And in a month, uh, where we are is we have um shipped in IT terms, we have shipped and deployed almost 70,000 lines of code. We have over 15 applications now that are live and working.

Future Implications for Nonprofits

Bill Simmons 14:53

The cost, if I had had to have this coded and produced, it would have cost over $750,000. And I'd still be waiting for it another four to six months to have it delivered. And so we have a whole build plan, and 80% of that code is code that I'm delivering, or clawed code is delivering, but I'm architecting and delivering. So that's an entirely inverted um way to work uh and exist in the world. And you know, it all goes back to this moment of seeing that anything you think of is possible, and then understanding that we're at a place where the tools can deliver real meaningful results inside an organization, and that changes everything.

Adam Jeske 15:41

Wow. Um I I talk to a lot of CEOs, and I have not heard anyone picking up the challenge or the opportunity the way that you are. Tell us more about what you have built, what problems you're solving using the new technology.

Bill Simmons 16:00

Yeah, it's basic uh problems. Uh, and it it's revealing that inside small and medium organizations, there's actually a backlog in software and automation. Um, because you can't typically afford the sort of software that really optimizes your organizational workflows or enables uh you to access data. Data tends to live, even if you have access to data, tends to live in a very lag state. And so decision making is, you know, every quarter you meet with a marketing firm, and so you then you get an update. Um, you know, your financial data, you get your your financials, but you know, being able to see patterns uh is a longer term uh question or depend is very dependent on the individual capability of any one team member. Um so you might have a great dashboardist in a in a CFO sort of role, um, or you may not, and then so then it's dependent upon the person. So we're delivering right now a lot of decision-making tools, and so people who are in fundraising, giving them uh completely different views, both of performance of of their portfolios, it's uh then push information. This is where you know you leverage data and you start prompting then your team, and so then you connect the whole stack together. So, you know, you get normally you might get a report, then someone might read it. Then you're like, oh now what do I need to do? Uh and then they put it on their to-do list. Well, if you can take the data and then you can have uh have AI then both present the dashboard, but then also prompt Adam, Jeski, all right, now here's the 10 people you need to call today, and then it creates the Asana task for you to do, and then the and you're all in the same UI, and then you're able to do the task, gets checked off, that flows back into the CRM. Everything is all inside one place, but it's data that was disparate or difficult to get to. And so uh, you know, you could play that sort of scene out in multiple ways right now inside our organization from uh traditional direct mail segmentation, which usually took a very expensive external agency to manage. Now we have all of that technology built inside the house. If someone needs an analysis on a report, we also have branded approaches and consistent approaches to report building. So, you know, one of the things that Claude can do, Claude uses something called skills. And so when you use Claude Teams, um, you can create a brand skill. And so when you do that, now any team member in our organization, whenever they create a document, whether it's Word, it's PowerPoint, doesn't matter what it is, the brand skill gets activated and it's immediately branded. So you never have to remember, never have to worry about brand conformity any longer. It's automated output inside it for every team member. And then to have that delivered in a way, so whether that's an internal report or it's a donor report or a board report, to have that brand cohesion uh in a brand change organization is really important. But to to have your team members remember, oh, don't forget to use the you know, the brand template for your letter head, or don't forget to use the PowerPoint template for your PowerPoint with the logos on it. I mean, all of this is a lot of human in the middle interference, and all of that's gone. So I'll give you a very tactical implication for us. So an organization our size is too uh big and too small. So a smaller organization probably couldn't afford an external marketing agency. Um, a larger organization probably wouldn't worry too much about having a marketing agency, it would just disappear under the overhead cost. We're in the middle spot where we sort of can afford and we sort of cannot afford to not have one. Sort of everything is true. And we're on a path right now where by July 1, the tools that we will have

Cost Reduction and Efficiency Gains

Bill Simmons 20:19

built will enable us to operate at 75% less of a cost, and I'll just use some marketing, external marketing consulting to keep a human, a really qualified human brain in the mix. Um, but we will reduce our cost, we will reduce our fundraising cost by 25%. Our fundraising budget will go down by 25% because of not spending it on a marketing firm. That's real bottom line impact uh that then goes back in, that then gets used to take advantage of the fact that now we have data that is closer to us and we can act on it immediately. And the sort of so driving, you can drive costs out of the system, and I don't I think that's that only lives so long. So it's not uh it's not, it's it's nice to have, but it's not the primary motivation. The real motivation is to give us the ability on to act on uh decision making quicker by putting the information in front of ourselves. Then I can redeploy that savings into ways that will generate more revenue for the organization. Uh, and we become more resilient then. Uh, we're not dependent on an outside agency. And, you know, if if I told my board or myself three years ago, I'm gonna bring all this in-house, they were, that's that's crazy. We can't do that. We don't have the people, we don't have the capacity. Now I do. We're almost done building the fundraising replacement technology inside the organization. And and Claude and I are about a month away from being finished with that whole project. And that will revolutionize. That's why it's a huge step because it will revolutionize how we operate as an organization. And move us into a whole new frontier of access to technology because the patterns that AI can see in the data, we can put so many more pieces together than we ever could. The marketing agency can only see certain data points. But we can put all the data points now in and look at things we've never looked at before and be able to send the Adam Jeskys of the world a much different sort of campaign and appeal or ask than we ever did before because we couldn't look at all the data at once. And AI in the middle helps us to do that. So yeah, it's um it's been a pretty crazy ride so far.

Integrating AI into Workflows

Adam Jeske 22:44

If I'm tracking correctly, there's obvious impacts around different facets of revenue generation, marketing side, uh as well as some of the major donor work. Are there other parts of uh you talked about the report generation, some of the task generation, getting things uh sketched out task-wise within Asana? Are there other significant parts of the organization that are being effective outside of those that I just ticked through? I might have missed some that you've already mentioned, but maybe there's others you haven't mentioned.

Bill Simmons 23:18

Yeah, I they all are. Um, and so one of the things let me back up and say so we use Claude teams across the organization. And then uh for a little over a month now, I've been doing training every Wednesday. I have a training this afternoon to walk the team through uh migrating fully to using Claude inside their workflow. How do they use it? How do they optimize it? How does it make their work better? So that's important. I don't need the whole organization to have the depth of knowledge on Claude Code that I do. I have another person who is using Claude Code, and I have one more person that's onboarding into Cloud Code. So we're using it to some extent. Um, but it's not important. The whole organization, and I wouldn't want, I couldn't, there's no way you can have the whole organization use cloud code. But I need everyone to use Claude and to integrate it in their workflows. So it's changing everyone's workflow, and because we can control things, like I mentioned, like the brand skills, so that means every every person's work is now going to be changed by the fact that we're using these things called skills. The next step is the frontier, is we did a whole, we called it a job reality check, and everyone went through their roles and we had an assessment tool that Claude helped us build, and they basically talked about at a very high level where their time goes and for those time blocks, how much of that is described as a process that's already documented or not. And then Claude helped us build a review of that so that we could understand across the organization, regardless of discipline, how much work could be automated to free up people's time and then where the gap was in documentation of process so that you you can't automate if you don't document. So we identified that in a small organization because we we work through partners around the world. So we don't have a huge team. Our team is between 17 and 20 people. And uh we could automate about three to four FTEs worth of time if we started to go through those steps. And so that informs a development pathway that we're on. So we're documenting and then automating. And so that will affect every workflow across the organization where there's something that is has the potential to be automated. It's on a timeline and a development path to be automated, and that's going to free up more time. I'm not trying to eliminate positions. We're a small organization. So I got people who need time back and need to do things more productive and make better decisions. And so that's the next. Then there's the big boxes of like actual software delivery because automating a process, I'm not really counting in development, even though it may take software to automate something. Then the layer above that is actual software tools that persist. And so something that could we have I developed an internal

Building Internal Software Solutions

Bill Simmons 26:24

tools portal basically that uh holds all the apps we developed. And so everyone in the organization goes there and they can launch the app and they could run the tool, whatever the tool is. And so that sits. And now here's the not only did I save money, Adam, but here's the other side of it. I did this inside our existing stack. And so we have had I've had no new software licenses, I've spent no new money. So all we use from end to end, we use we use Google Workspace and Salesforce and Sage for our accounting uh and Asana. Uh, and so everything is developed inside that. I'm not using a solution uh external to our existing stack. So I've had no new, I haven't spent any more money on software to do this other than my own license for Claude at the high level. Um, but other than that, we have no new software in the organization. And so we're we're not only developing over $700,000 worth of software, but we've done it for no real um increment. We now we have we're we are tracking our internal cost of development. So we've invested my time you know is real, right? And is finite. And so, you know, we've invested about $25,000 of staff time to deliver that $800,000 or so of value. But my development cycle will taper off. We're not a software development company, so I'm not trying to produce more products to sell into the marketplace. Although I I do I do have an idea that I think I have several ideas I think that could generate revenue um because we're solving problems that small and medium-sized nonprofits have and uh with long-term software solutions. And so I'd end all of that by saying I think that most people think about AI today as this, you know, I think 92% of organizations are using AI, 80% don't have any documentation, there's no real method to why how it's used, and it tends to be these chatbot sort of interfaces. And so I don't think about AI that way. I think about AI as using tools like Claude Code to transform organizational work and to do things we could never have done. Now it takes someone who is a system architect who can think that way to do it, but I think it's an important question. If I invert the question, if I invert what we're doing around, I had this um uh Albert Chen and I had this conversation recently. He's uh somebody we mutually know who is in the in the AI space, but he and I were talking about what does an AI native nonprofit organization look like? If you're gonna start an organization today, what would you do? Who would you hire? Uh and I think um I mean my answer to that question is very different because I'm pushing an organization down to AI efficiency using AI tools.

The Future of AI in Nonprofits

Bill Simmons 29:16

But if I was starting one tomorrow, it would be very different. I I could tell you the first uh one or two people I would hire. I would hire two people who know how to use Cloud Code before I hired anybody else. And then what I couldn't build, then I'd eventually I'd hire for. But I that's that's what I would do if I was starting an organization today. It's too possible to build for free. I mean, the a great example is you know, if you ask Cloud Code to build a user interface that operates an operating system that looks like uh Apple iOS, you know, it'll do that in under half an hour.

Adam Jeske 29:49

Yeah.

Bill Simmons 29:50

You know, how much money do you think Apple put into developing, you know, what we're used to in iOS? I I mean, you know, that's that's the sort of thing it can do. And when you can direct that sort of power at a very specific problem that you have as an organization, you can get a solution. You just have to ask it. Do you have to declare the problem so it can help you find the solution?

Adam Jeske 30:11

Yeah, sometimes that assessment or or definition, where to point all that new power is really the core question.

Bill Simmons 30:21

Yeah, I that's you know, I the analogy I've been using is that, you know, when electricity was invented, most people took the candle on their table and they just replaced it with a light bulb. So nothing in the room changed. They're just now it's a bright, it's a brighter room. And I think that's the way most people interact with AI today. You know, it's the same work, the room's a little brighter, maybe they're getting a little thing done here or there. What we're doing is we're basically wiring the building and putting outlets in and building appliances to run on the outlets. To me, that's a completely different use uh for the tools that are available today. And most people can't see that far because they're the way they onboard it to these products is through a chat interface, which is a very ineffective way to share the capacity of the technology. And so people now are sort of patterned into this condition around how they interact with the thing that is actually not its core capacity. Uh so people have mistaken it for a light bulb, and it's it's it's much, much more than that. Great metaphor.

Adam Jeske 31:27

Did you come up with that, or is that from someone else? That's Claude, yeah. Claude came up with that one for me. So great. Um, it's a really good picture. So uh I'm I'm watching our time and I want to get through a a couple of additional questions. Uh, you mentioned uh Albert Chen, our mutual friend. Uh, he runs anago.ai, a n-a-g-o.ai, really thoughtful person who helps organizations on this front. Uh aside from conversations with Albert and just trying stuff for CEOs who are like hearing this and thinking, oh, I need to start doing some stuff. What do you recommend to them? How do they get going? How do they learn? How do they stay abreast?

Onboarding and Training for AI Tools

Bill Simmons 32:16

Yeah, that's a great question. I I did do a workshop for a fellow CEO in uh the Accord Networkspace. And but he already had uh so you know some background awareness of coding, and they had a big tool friction in their organization. And so I did a training for he and and a few of his staff. And and just a few weeks later, he had a full running uh application and removed that friction and and onboarded pretty quickly. And I basically just handed this some onboarding tools. And so I I there's sort of two ways to get to this. I'm talking about uh onboarding to a high-level use of this, I'm not talking about sort of someone who's just trying to learn how to use AI. That's what I'm training my team on. That's a whole different sort of a training. But if someone's interested in uh they think, you know, I think I could figure this out. Because I understand that's a small audience, not everybody, you know, is needs cloud code, would understand how to use cloud code, understand all the architecting thing decisions and how to keep it all safe, right? It's another question. Is all this secure? Is it, you know, am I following good protocols for donor data? You know, how is all that managed? So if you can't visualize how to do that or to get cloud to help you do that, then you create risks that you don't want your organization to have. There are some tools out there that make using things like Cloud Code easier today. Uh Albert Chan and I both have another friend, Carl Wirth. He has a product that's called Nimbalist, N-I-M-B-A-L-Y-S-T, Nimbalist. And Nimbalist sits on top of Cloud Code because Cloud Code looks like you know a 1980 computer interface. Um, it sits in terminal, and so it's not very interesting if you're not into coding. Nimbalist puts a user interface on top of that and makes it very easy to then interact because Cloud Code, it will accept a coder being in front of it, but it'll also accept you or me and just using a regular language, plain language into the prompt. But Nimbalus makes that much more simple and has radically facilitated the speed of development that we've been under. It's a free product still because it's in alpha or beta. Um, so I'm not I'm not selling something that uh is going to cost people money. But I would that's what I would do. I would on-ramp, and I would I would have Claude train you how to use it. I mean, I that's you declare the problem to Claude in Claude Code and in Nimbleist if you want to, and say, how do you do this? Because here's what I do. Here's my tip for a developer, here's my developer tip. All right. Uh here's my discipline. I use Claude.ai in a large project I built. I use the regular Claude chat interface for architecture. And so it has a whole skill and memory that I've taught it that says it never codes, it never, because it will want to code, it will want to create the solution itself. But it knows, nope, I'm the architect. That's all I do. I'm the architect. So I have a window open with Claude AI where I'm talking about the problem. We start design the software we need, it understands all the IT stack that we have, and then it starts to propose the solution. Once we've iterated that enough and I like it, then we hand that markdown file that it generates. I hand that to my Cloud code and Nimblest, and it goes off and it starts developing. And we use a phase development process where it basically says, all right, do faith, don't do all this at once, because you could probably do that. Let's put some interruption in here and some friction and make sure everything works as we go so that by the time we're done, we can publish running software. Now that's a software development practice, uh, but the way I got there, the way I learned was by using regular Claude, interacting with it around what I wanted to build, then letting Cloud Code go do and actually build the thing. Is it more complicated than that? Yeah, sure, there's more things that have to be set up to actually run Claude Code and to store code and to make it work and to automate, you know, my Gmail, the agent I built that runs my inbox and filters and archives and trashes and creates drafts. You know, it has to do about seven different things to actually do that every night. But I just had to ask Claude how to do it. And here's the number one lesson that I had in my in my Claude learning, my Claude education. Um Claude will try to present solutions, and most AI probably will try to present solutions to your problem. And we'll do one of two things. They'll either say, Here's, oh, there's a um there's a software product called Zapier that you can use and you can go pay for it, and it will do, it'll hook all these things together. Um, that so uh a software solution is one thing it may do. The other thing it may do is it may say, Hey, can you go into this tab and click this box? Neither of those things uh need be possible. So when Claude suggests Zapier to me as a solution, I simply ask it, can't we code our own solution for that? And it'll say, Oh yeah, sure we can. And there's zero cost in that, so let's go do that. We'll build that and it'll have a zero cost solution that's that does what I would have paid Zapier for. Right. I also have taught it when it has me opening a window, it knows to never ask me to do anything. Once you've given it permission inside your software, so it can get inside anything that's Google, anything that's Asana, it can get into anything that I that I can get into. Uh every now and then it'll forget and it'll say, Hey, can you go open this and uh and click this? And I'll say, No, I'm not doing that. You can do it. And then Claude says, Oh, yeah, okay, I I can't do that. These are my small friction points with with Claude. But it's a different way to go back to my very first point, and that is anything that you can think of is possible. So it may suggest a software solution. A software solution is software, it's code, and Claude Code can code code. And so, and and then anything that I can do on a computer is just me interfacing with code, which means I can tell Claude, no, I'm not going to do that. You do that. That involves clicking on a computer. Uh there there are like one out of a hundred times where you actually have to hit authorize and and Google. Okay, that's fine. I'll accept that one friction point. Anything you can think of as possible and onboarding to it, the tool to onboard yourself to these tools, if you have the imagination for it. And that's the word we use in our organization: imagination. I want people who can use their imaginations because anything they can imagine inside the workplace, Claude can help us build. And that's the friction point. It's not training you how to use AI or tools. Imagine something, talk to Claude about it, and try to force Claude to do it. And let Claude know that it has a development partner called Claude Code that can do the coding. And you might be surprised what you get out of it at what it can build.

Adam Jeske 39:27

Uh Bill Simmons, president and CEO at Hope Rises International. What do you love about your work?

Passion for Problem-Solving in Nonprofits

Bill Simmons 39:34

You know what? I I I told my board chair just this week, I said, you know, I am I love this right now. I love that I can deliver solutions to my team. I mean, there is nothing like giving someone who's on the phone with a donor and it has friction. I mean, you know, this is, you know, and Gallup's Q12, you know, having the tools and resources you need to do your job is one of the 10 questions. And uh, and to be able to just immediately now provide a solution and a tool for people that helps then facilitate our donors doing what they want to do in the world or help us accomplish our mission, it is really exciting. It sounds may sound strange, but man, I'm so engaged and excited about this moment, which is uh equal parts terrifying uh and equal parts fun. But that's it just sounds like a roller coaster. And then, you know, why'd you get on the roller coaster in the first place? And so I'm loving it.

Adam Jeske 40:30

Man, um, it is rare, Bill, to speak to a person of your tenure, 15 years or so, um, who has this sort of energy and innovation, and also to be sort of the CEO and what it sounds like sort of this chief uh artificial intelligence officer as well, which some places are actually implementing and hiring in uh larger organizations. I have so many additional questions, like how is this affecting how you lead your executive team? Was your tech stack already defined as you started the journey, or did you have to make some choices about that? How do you uh knowing you a little bit, there's some uh organizational discipline that I think is how you're wired and how you probably were leading before this point, but I also see that lacking in many organizations and many CEOs have not built uh to a point where they would be ready to continue building uh using the the new tools. So that would be interesting terrain. I think uh I would like to talk again at some point. Yeah. Thank you for joining us on the nonprofit CEO podcast. Yeah, well, thanks for having me, and I'll I'll look forward to another conversation. It'd be great. If you know a nonprofit CEO who's carrying something significant, send them the show. And if you want more of the patterns I'm seeing across hundreds of CEO conversations like this, you need the nonprofit CEO briefing delivered weekly at nonprofitceo.com. I'm Adam Jeski, the Nonprofit CEO Advisor. Thank you for what you're leading.