AI instruction for development and program staff
Twenty years in nonprofits.
A business run on AI.
Noah Gallagher directed summer camps for twenty years and trained teams of managers, at every age, every one of those years. In 2026 he opened a photography and consulting business and ran it on AI from the first day. The courses teach what he uses to the people who run the annual fund, the programs and the office.
Live, in groups of eight or fewer. Six in train-the-trainer. Massachusetts employers may be reimbursed.
What he built with it.
Nothing here needed a programmer. Each piece is a task an office already has, done with AI and checked by a person.
A page for every client that updates as the work moves
Each engagement has its own web page: the plan, the photographs as they are finished, the guide to using them. The client's team reads the page instead of an email thread. One of them, for a camp in Wisconsin.
A proposal to a philanthropy director
The store with a share to the camp, the donor and board book, framed limited editions for the auction, stewardship gifts, and the whole library curated. Beside the framed-editions line, the piece itself: a photograph that becomes a twenty-two second film of itself on a wall, rendered on his own computer for nothing.
A business that runs itself around him
At 7:30 each morning a routine reads the mail and the calendar, writes the day's plan, and asks him the questions the record cannot answer. His answers go back into the record the same day. Before anything reaches a client, three other AI models read it, for the buyer and for the facts, and each finding points at the line it means. A memory carries across sessions, one fact per file, so a mistake he has caught once is not made twice.
Each course teaches one piece of this, sized for an office that is not his.
What you practice, by name.
The field, this month: the major nonprofit CRMs now draft acknowledgments, appeals and stewardship updates and score donors. Funders have begun asking whether AI helped write the proposal. In the NTEN and Bridgespan survey of September 10, 98% of nonprofits use AI, half in fundraising, and 32% of staff have had formal training. The gap is the person.
- A year-end appeal and the acknowledgment letters that follow it, drafted from your case for support and checked against it.
- A stewardship update or impact report for a mid-level donor, built from program facts, with nothing invented.
- A grant report or proposal draft from your own source material, with an answer ready for the funder's AI-disclosure question.
- A context document that gives AI your voice and your program facts while gift amounts, constituent records and donor lists stay in your CRM.
- A saved instruction set your annual fund team reuses every cycle, handed to a colleague with a way to check their work.
- Teaching one of these to the person your team already asks.
Most teams start with one of the first two courses. Tell us the task and we will say which.
Tuition is per participant. Four enrolled participants are needed for a cohort to run. Organizations may send individuals into a shared cohort. Software subscriptions are additional. Open a course to see what you practice and leave with. The smaller line under each name is the course's registered title with the Massachusetts Express Program.
Train the trainers
Give your team someone
who can teach this.
For the staff member who runs onboarding, supports colleagues, or is already the person everyone asks about AI. Two thirds of nonprofit staff have had no formal AI training; this course makes one of yours the trainer.
Each participant builds and tests a reusable AI skill, turns it into a lesson, teaches it to a peer, and watches that peer use it on a new example. Feedback covers the skill and the teaching. This is the work Noah did for twenty years: training the managers who trained everyone else.
You leave with the lesson plan, the practice exercise and a way to tell whether a colleague can do the task alone.
Twelve hours of live instruction across four sessions. Six participants, so every person teaches and is observed. $1,800 per participant. Prior AI fluency required.
Try the method
in one minute.
Three notes about a staff orientation. First the notes, then what an AI drafted from them, then the draft after a person checked it. The skill is the third step: seeing what the source never said.
How a course runs.
Live, on a video call, during paid work time, in groups of eight or fewer. Two to four sessions of two or three hours, a week apart, so what you practice in one session gets used at work before the next.
Each session has the same shape: a task from a fictional nonprofit, your first attempt, the check against the source, then a new case on your own. Trainers also teach the skill to a colleague. What you can do alone is recorded; attendance is recorded separately. The standard throughout is the one the field already holds: a person reviews every AI-touched piece before a donor sees it, and constituent data never enters a public tool.
Massachusetts training funding
Massachusetts employers with 100 or fewer employees can be reimbursed up to 100% of tuition through the Express Program, up to $15,000 a year. Eligibility depends on how your organization pays unemployment insurance: nonprofits on the contributory method qualify, those on the reimbursable method do not, and government agencies are not eligible. We help you check before you apply.
Our first course is under review with the program; the others follow that review. Reimbursement depends on course approval, your eligibility and an approved award. An application must go in at least 21 days before training starts, and training does not begin before the agreement's start date; we walk you through the steps. You can inquire about any course now.
What would you like your team to do?
Tell us about one task and who needs to learn it.