AI Consulting & Strategy
Claude
Consultants: Pallari Mujumdar
Since 1973, Mattie N. Dixon Community Cupboard in Ambler, PA has lived our its “neighbors helping neighbors” mission, providing, food, clothing and support to individual and families with incomes within 150% of the federal poverty guidelines. Grants are essential to funding that work, but finding and applying for them consumed 10-15 hours a week during grant season for its Executive Director, Cindy Wedholm, who manages the full grant cycle without a dedicated grant writer. That time was pulled directly from programming and day-to-day operations. The organization partnered with Green Leaf to design an AI-powered workflow that could streamline the grant process while preserving the Executive Director’s voice and control over every submission.
With a small staff and no dedicated grant writer, the Executive Director was spending 10-15 hours a week during grant season searching for opportunities and manually drafting proposals, time pulled directly from programming and operations.
Green Leaf built a Claude-powered agent workflow that searches for grant opportunities, rates each one’s approval likelihood, and drafts proposals and letters of intent (LOIs) in the Executive Director’s own voice, with a built-in human review step before anything is submitted.
Grant funding is a lifeline for Mattie Dixon, but the process of finding and winning it was consuming time the organization didn’t have to spare. As Executive Director, Cindy Wedholm wears many hats, and grant writing was one of the most demanding.
Every cycle started the same way: searching foundation websites, filtering opportunities by fit and geography, then working through each application by hand, copying questions into Word, drafting answers, and pulling budget figures for financial reporting. She tracked it all in a spreadsheet: what she’d applied for, what was requested, what was granted, and when each report was due.
During peak grant season, that added about 10-15 hours a week, on top of running the organization. The work was especially draining on general operating grants, where application questions were often written for project-based funding and didn’t map well to the kind of ongoing, wholehearted answers she wanted to give. Every hour spent on applications was an hour she couldn’t spend on programming needs or the organization’s day-to-day operations, and she has to lean on a small staff to cover the gap.
Green Leaf designed a Claude-based agent workflow to take on the repetitive, time-intensive parts of the grant process while keeping Cindy firmly in control of what gets submitted.
Her biggest concern going in was that AI-written grants would come across as robotic. So before Claude ever searched for an opportunity, Green Leaf fed it 10-12 of Cindy’s past approved proposals to build a “Voice Profile,” a reference to her tone and sentence structure that keeps every draft sounding like her rather than a generic AI.
From there, the workflow fits naturally into her existing routine. Cindy logs into the grant portal herself, then simply prompts Claude to pull a list of prospects. Claude does the research, weighing each funder’s criteria against the organization’s own profile, and comes back with a ranked shortlist and an estimated approval likelihood for each one. she picks a prosect, reviews the draft Claude produces, and can even have Claude fill out the funder’s online application directly.
What makes the system trustworthy is that a human stays in the loop at every stage rather than being asked to sign off at the end. Claude never submits an application on its own, and it flags any figures or details that need a second look before handing a draft back for final review. That same instinct shaped how the workflow was refined: when early testing turned up prospects that tool has already surfaced, Green Leaf added duplicate-checking to catch repeats, and when broad web searches returned weaker matches than the grant platform Cindy already trusted, the search was narrowed to that trusted platform alone.
Two days after seeing the tool in action, Cindy’s reaction was amazement, particularly at how well it reflected her own voice and pulled context from past grants automatically. While she’s still building hands-on comfort with the workflow, she’s already expecting real, measurable change in how she works.
Based on what she’s seen, she anticipates her weekly time on grant writing will drop from 10-15 hours to roughly 5, and that freed-up time will go toward programming, operations planning, and thinking ahead about how the organization can keep growing to serve its community. With less time lost to the search-and-write cycle, she also expected to be able to review and pursue more grant opportunities than before, right as she heads into her busiest season of the year.
Discover how Green Leaf Consulting Group’s strategic IT solutions can bring you growth through technology. Contact Neil McCole at neilm@greenleafgrp.com or (484) 535-4405.