(TL;DR – Use the new AI Grant Helper to build professional grant prompts for AI.)
If you’re a fundraiser, you’re probably doing one of two things with AI right now. You’re either ignoring it entirely — too busy, too skeptical, too overwhelmed — or you’re typing something like “Write me a grant narrative for a food bank” into ChatGPT and getting disappointed by the result.
Both responses are completely rational. If you’re in the first group, you don’t have time to experiment. If you’re in the second, you experimented and concluded the tool doesn’t work. I understand — when I tried it six months ago, my results were garbage too.
But there’s a third option.
For smaller agencies, AI could be a game changer by helping them put together compelling grant content. But getting a polished result requires something most fundraisers don’t have time to figure out: a really good prompt.
How I Got Here
In January, I decided to lean into AI in a new way. I’d played around with ChatGPT and image generation and thought it was fun but limited. Then I discovered 100school’s 30 Days of AI challenge. The headline was “Upgrade your professional operating system, rewire how you think, work & create with AI.”
I gave it a shot, and it blew my expectations out of the water. Every day I received an email connecting me to a lesson that introduced new ways to think about AI, as well as some of the top new AI tools, like Google’s NotebookLM, website and app builder Loveable, and Claude, the AI assistant from Anthropic.
The real game changer was Claude Code, Anthropic’s developer tool. Imagine having an expert coder working alongside you. I’ve been doing web development on the side since 2018, and suddenly here was a tool that could automate 90% of the grind work. The idea and the vision were mine, but Claude handled the grunt work. In the course of two weeks, I built a video game, YourFirstBillion.com, and pushed it to the web. My 13-year-old son is now working to improve the game and make it more fun.
When I finished the challenge, my thoughts turned to fundraising, and in particular to grant writing. I know that there are a lot of ministries that can’t afford a professional grant writer, but are still doing great work — work that could get funded if they were able to put their efforts down in words.
The Prompting Problem
Research from Anthropic suggests a strong correlation between the quality of the prompt you submit and the quality of the output you get back. A question that sounds like it was written by a Stanford professor will get an answer that sounds like a Stanford professor, while a prompt that sounds like a middle schooler will get an answer that matches.
So to translate to grant writing: a prompt that comes from a person who doesn’t understand food banking produces content that sounds like it was written by someone who’s never set foot in a food pantry.
The problem isn’t that you lack the knowledge. It’s that nobody has shown you how to translate that knowledge into language an AI can actually use.
What If You Had a Tool That Wrote the Prompt for You?
The idea behind my new AI Grant Helper is simple: what if, instead of asking you to become a prompt engineer, we asked you the questions you already know how to answer?
You know your mission statement. You know who you serve. You know what you’re asking for and why. You know whether you’re applying to a community foundation or a federal agency. That’s not new information — it’s the stuff you live and breathe every day.
So I built a tool that translates for you. You answer 17 questions about your organization and your grant. The tool assembles those answers into a detailed, sector-specific prompt — complete with research instructions and tone calibration — that you can paste directly into Claude or Gemini.
How It Actually Works
The experience is intentionally simple. You land on the page and see a single question: What’s your organization’s name? Answer it, and the next question appears. Then the next. One at a time, so you’re never staring at a wall of form fields wondering where to start.
As you answer, a live preview builds on the right side of the screen. You can watch your prompt taking shape in real time. There’s something satisfying about seeing your mission statement, your service area, your funding request all woven into a professional prompt that you didn’t have to write yourself.
The questions move through four sections: your organization, the grant you’re applying for, the specific application questions (or the types of content you need), and any supporting data you want to include. A progress bar tracks where you are. When you finish, a modal pops up with the complete prompt, ready to copy or have emailed to you.
That’s the user experience. Simple, guided, no account required.
The Part That Took the Most Thinking
Here’s what I didn’t want to build: a generic prompt generator. “Write a grant for [organization name] requesting [amount] from [funder].” That’s not useful because it doesn’t understand that a food bank grant narrative needs to sound fundamentally different from an arts organization grant narrative.
So the tool calibrates. When you select “food bank” as your organization type, the generated prompt tells the AI to write with a data-driven, community-focused tone. It instructs Claude to research food insecurity trends in your service area, to look up your funder’s stated priorities, and to align the narrative accordingly. Select “arts organization” instead, and the tone shifts to cultural impact language. The research instructions change.
This is the part that draws on actual grant writing experience. I’ve written hundreds of successful grant proposals, and different grantors are looking for different kinds of information. A government grant will have very different questions from a private foundation. Both will be different from a corporate foundation. Getting the tone and the content right isn’t a nice-to-have. It’s the difference between fully-funded and the “circular file.”
The other design decision I’m proud of is what I call the “research step.” Most people treat AI like a typewriter — type in a request, get text back. But Claude and ChatGPT are capable of much more than that. The prompts this tool generates include explicit instructions for the AI to research your sector, your funder, and your community before it starts writing. It tells the AI to look up recent trends, find relevant statistics, and understand the funder’s giving patterns. All before it starts responding.
That research step is the difference between AI-as-shortcut and AI-as-consultant. A shortcut gives you faster mediocrity. A consultant gives you something informed and tailored.
What This Is Really About
I’ve been a fundraiser for over 14 years. I’ve written more grants than I can count. And I’ve watched this profession get harder — more competition for fewer dollars, more reporting requirements, more complexity — while the tools available to most fundraisers haven’t changed much since I started. Word docs and Excel spreadsheets.
AI is the first genuinely new capability to enter this field in a long time. But it’s only useful if you know how to use it. It’s not just for the organizations with six-figure tech budgets and dedicated innovation teams. It’s for the leader at a small food pantry who’s writing grants at 9 PM after getting home from their day job. Or the one-man show at a rural arts nonprofit who does everything from major gifts to the newsletter.
That’s who I built this for. The prompt is the bottleneck between you and the AI tools that could genuinely help you. Remove the bottleneck, and you tap a resource that can help put money in your bank account.
If that sounds like you — head over to The Almoner and build your first prompt. Ten minutes, and you’ll walk away with something you can use tonight.
