AI Studio
Let's Talk
Stacks AI Studio / Small Nonprofits
"I feel behind on AI, and there is no chance I can catch up." If that is the sentence in your head, you are exactly who I work with. I am not selling software or theory-based advice. I run my own ten person managed services business on Claude, I know the value it adds every week, and now I help owners do the same inside small nonprofits like yours: your team trained, jobs automated, hours saved. Saying yes to starting puts you ahead of many of your competitors.
Experience, Not Theory
That is our first four automations at Stacks, across a ten person team. We took on more clients, and gross margin went from 26% to 56% in twelve months. AI was not the only driver. It was the biggest one. It also set a new standard for what my team sends out: the same bar for research, formatting, and finish on everything that leaves the building, no shortcuts.
Casey Templeton, Founder of Stacks. Nashville.
"In one week I went from a two-hour session with Casey to a brand new marketing site and firing the marketing firm I'd used for two years. If Casey wants to work with you, don't be a dummy. Just do it."
The Week
Five jobs that land on one desk. This is the work an automation goes after first.
The report is due and the numbers are scattered. Somebody pulls program data out of three spreadsheets.
The LOI lands and a full proposal is due. The last proposal gets rebuilt line by line.
Gifts arrive faster than the thank you letters. A first time donor waits two weeks to hear anything.
The board packet gets built the night before. The program numbers and the finance numbers never match.
The year end appeal starts from a blank page. Last year's letter is buried in somebody's drive.
The grant report, the acknowledgments, and the board packet all land on the same desk.
Source: Center for Effective Philanthropy, May 2026.
The First Step
For most small nonprofits, this is the first step. Homework first, then two sessions, two hours each. In person if you are near Nashville, or over Zoom. $1,950.
For the math: one person at $25 an hour who gets two hours a week back is worth about $2,600 a year. If the first automation does that for one employee, it pays for itself with room to spare. And if after our first session you do not see a job worth automating, you pay nothing.
An automation is a repeated piece of work you teach Claude once. Your team runs it every day, you own it and tune it, and it carries the work to 80 percent so your people finish the 20 that needs their judgment.
"Casey taught me how to actually put AI to work. I generate quotes faster, my workflow is smoother, and I'm making more money because of it. One of the best investments I've made in my business."
01
Before we ever meet, you pull real examples of the work that eats your week: actual documents, actual messages, nothing cleaned up for me.
02
Two hours. We map how your organization actually runs, get Claude set up for your team, and train you on how to operate it.
03
Two hours. We start building your first automations and put together the training your team needs to run Claude every day.
04
Claude running for your team, your people trained on it, and one job automated and in use.
One page, no obligation. Proof of what we could build together, before we ever talk.
One email with the PDF. No newsletter, no drip campaign.
of US small businesses say they use generative AI.
Source: U.S. Chamber of Commerce, August 2025.
Every application and report goes out under your organization's name, so the executive director reads and signs each one.
The Stack
It works with the tools you already run. It reads from your donor database, your grant calendar, your program records, and your books, and writes back where it can.
Start with one connection, not all of them. The first automation usually needs one export from your donor database, and that is the whole first step. Once it earns its keep, the next system comes in, then the one after that. Every automation gets sharper as it goes, because Claude can finally see the whole organization instead of one screen at a time.
The Numbers
That feeling of being behind is not in your head. Small nonprofits are adopting fast, and the money side got harder at the same time.
Social impact professionals surveyed using AI
Source: Blackbaud Institute, June 2026.
CEOs say foundation grants got harder
Source: Center for Effective Philanthropy, May 2026.
Questions
No, and you are not as far behind as you think. The owners who feel furthest behind often move fastest once they start, because they have not spent a year on tools that did not stick. You do not need a strategy, a budget cycle, or a technical person on staff. You need four hours and one job that is costing you time every week. That is the whole first step.
No, and that is the whole design. An automation carries a job to 80 percent quickly and your staff takes the final 20 percent over the finish line. Efficiency goes up for everyone, the organization grows without the headcount it used to need, margins grow with it, and your people spend their time on donors and program work instead of retyping.
Usually, whether that is Bloomerang, Little Green Light, DonorPerfect, or Salesforce. Some systems give me an API, some give me an export, and a few give me nothing but a screen to copy from. I tell you which one you have before I scope anything.
No. Everything I build is admin. An automation drafts, summarizes, and organizes. It never scores an applicant, ranks a family on a waitlist, or decides who receives services. Those calls stay with your program staff.
You approve the data path before any automation runs. Client records can be as sensitive as anything in health care, and I treat them that way. Where a job can run on redacted material, it runs that way. Where it cannot, the path is written down and you sign off before an automation touches anything real.
Yes, as a draft, and you still sign it. The application and the report are attestations with your name on them, so an automation hands the executive director a draft built from your own program data. Some funders have their own rules about AI, and I build inside whatever yours say. Nothing reaches a program officer that a human has not read.
The first automation runs in weeks, not quarters. The two sessions scope it, I build it against your real reports, and your team uses it while I refine it. One automation in production beats five in a plan.
You own the automations and the documentation. They live in your account, written in plain language, so your staff can read them and change them. If you stop working with me, nothing walks out the door.
Then you pay nothing. If after our first session you do not see a job worth automating, the engagement ends there and there is no charge. The first session either surfaces a job worth the money or it does not, and that risk is mine, not yours.
Because the format is small on purpose. Homework, two working sessions of two hours each, and the setup work around them. For the math: one person at $25 an hour who gets two hours a week back is worth about $2,600 a year. If the first automation does that for one employee, it has paid for itself with room to spare. Every number in that sentence is yours to check.
Who You Will Be Working With
Founder of Stacks, a ten person business in Nashville that runs on Claude. Both sessions are with him directly, and every message from this page lands in his inbox, not a queue.
Start Here
Tell me about your nonprofit. I take on two or three of these engagements a month. Every message comes to me and I answer it myself. The first conversation costs nothing.
Book A 15 Minute Call"A fear of not understanding AI had kept me from diving in. Casey's coaching changed that. It opened my eyes to a sea of possibilities and left me more confident to explore. Come open-minded, and see where your imagination takes you."
I reply to every message. No newsletter, no drip campaign.