I had the chance to share the stage at the Northeast Annual Giving Conference (NEAGC) with Sean Devendorf, Executive Director of Alumni Engagement and Annual Giving at Tufts University. Together, we tackled a timely topic: how generative AI is transforming advancement work — not in the distant future, but right now.
Sean leads a high-performing, 30-person team. Over the past year, he and his colleagues were learning how to use AI tools like CoPilot or ChatGPT. “It felt like cheating,” they said. And honestly? That’s a pretty common sentiment.
But they worked to reframe AI not as a shortcut but as a way to enhance creativity, save time, and build more personal relationships with donors, volunteers, and alumni.
That’s how every leader should approach these tools. When used correctly, your LLM of choice can become you best friend — enhancing productivity and eliminating manual work so you can spend more time in face-to-face donor conversations.
🚀 The Opportunity (and the Gap)
Across advancement shops, AI is already being used for donor insights, predictive modeling, and automation.
Annual Giving and engagement teams are strapped for time and tasked with high-volume work — drafting emails, segmenting appeals or invites, and following up with hundreds (or thousands) of donors. This is exactly where generative AI can help.
Sean described how he encourages his team to integrate AI into everyday workflows, from planning campaigns to writing copy and analyzing data. His approach? Start small, show value quickly, and make the tools feel approachable — not intimidating.
💡Ideas to Get Started
Sean shared ways that the Tufts team is bringing AI into their day-to-day work.
- Hiring – CoPilot (Tufts’ approved LLM) helps the team outline job descriptions, generate interview questions, and summarize candidate feedback.
- Presentations – ChatGPT’s baked-in image generator, DALL-e, can help create visuals for internal presentations or social media posts. (Watch out, though. It’s not always great at spelling!)
- Analyzing Data – As Tufts builds a newly aligned engagement and annual giving team, they looked at giving society levels at peer institutions to generate new ideas for Tufts.
- High-level Planning – The Tufts team uses AI to map out complicated, long-term projects. Generative AI is great at quickly building timelines and suggesting deliverables to get big projects done.
Donor Experience Officers – DXOs at Tufts use AI to review prospect profiles, draft personalized emails, and write contact reports as they manage 1,000-person portfolios.
🧠 Better Prompts, Better Results
Pro tip: the better your prompt, the better results you’ll get.
If you’ve used Gemini or ChatGPT before but stopped using these tools because you didn’t like the responses, I encourage you to try again.
Not only are these platforms rapidly improving, but we learn more daily about how to generate the best results. And it all starts with giving more detailed, thorough instructions.
Here’s a simple framework you can follow to craft better prompts:
- Role: Who are you, or who should the AI pretend to be?
- Results: What’s the product you want AI to create?
- Context: What information does the AI need to know to create better responses?
- Intent: What goal are you working toward?
- Constraint: What are the limits in terms of length, voice, or medium?
Here’s what that looks like in action:
ROLE
I’m a Donor Experience Officer at Tufts University. I have 500 donors in my portfolio that have never met with a Tufts fundraiser before.
RESULT
Please give me a 7-step communication outreach cadence I can follow to book a meeting with these prospective donors. Two of the touchpoints should be via social media.
CONTEXT
Here’s a series of stories about the Tufts Annual Fund and its impact. Please pull from the themes contained here to create these touchpoints: https://alumniandfriends.tufts.edu/giving-to-tufts/tufts-fund-arts-sciences-engineering
I’m also uploading the latest issue of Jumbo Impact — a stewardship piece that illustrates the importance of giving to Tufts.
INTENT
My goal is to book a visit via Zoom or a schedule a phone call with these donors to understand their affinity for the university and capacity and willingness to make a leadership gift.
CONSTRAINT
Each step in the touchpoint plan should be an email message or video script that’s under 150 words and includes a different key element of the priorities and impact stories I’ve supplied via the link and pdf.
If you want to learn more about crafting better prompts or how AI and automation are changing fundraising, we have a free class for you: Modernizing the Donor Experience.
🔧 From Pilot to Practice
So, what does it take to go from theory to action? At Tufts, Sean has focused on four key areas:
- Leadership Modeling: He uses AI himself, shares examples with his team, and has created ongoing check-ins for the team to share successful AI use.
- Low-Stakes Practice: They’ve created space to “play” with AI, try out prompts, and learn together.
- Clear Guiderails: Every institution has its own rules hiding AI use. At Tufts, staff can use approved tools like CoPilot and Signal. They don’t use unapproved platforms for work.
The result? A culture shift — from skepticism to curiosity and confidence.
📬 What’s Next?
Whether you’re starting to experiment or already building AI into your advancement strategy, the bottom line is this: AI enhances relationship-driven work.
If you’re looking to get started, focus on your pain points. What takes up too much time? What feels repetitive? Try writing a prompt using the Role, Results, Context, Intent, and Constraint framework. Test it. Tweak it. Share it.
And most of all — don’t be afraid to experiment. That’s how we grow.
You’d better believe I used generative AI to write this. Here’s the prompt I used to create the rough draft.