Claude Cowork for Nonprofits
Move from chatting with AI to working alongside it. A five-week cohort for nonprofit leaders ready to put Claude Cowork to work.
- StartsJuly 30, 2026
- Length5 weeks · 90 min/week
- TimeThursdays · 11:00am to 12:30pm Central
- WhereVirtual (Zoom)
- Price$750per person
-
Ted Kriwiel -
Lucy Petroucheva
About this cohort
Tools like Claude Cowork represent a new generation of AI — not a chatbot you type questions into, but a system that can see your files, connect to your data, and take action on your behalf.
For the first time, nonprofits have a tool that can genuinely keep up with the complexity of their work, think alongside them, work alongside them, and truly transform how their work gets done.
But AI is only as useful as the context it has access to and the quality of the direction it receives.
This cohort is designed to help you build that bridge: connecting AI to the files, systems, and organizational knowledge where your work actually lives, while also learning how to guide AI effectively through strong prompting, thoughtful workflow design, and clear decision-making structures.
Together, we'll explore how you can create the "knowledge architecture" that makes AI genuinely useful in practice — from file organization and structured data to APIs, MCPs, and repeatable workflows that produce consistent, trustworthy results.
Along the way, you'll receive a practical crash course in Claude Cowork (released in January 2026). While we'll use Claude Cowork as the playground for the cohort, the principles apply broadly to working effectively with modern AI systems and Large Language Models more generally.
What you'll learn
In this cohort, you'll learn how to:
- Set up your knowledge architecture so AI can work across your files and tools
- Connect AI to the tools and data in your tech stack using APIs and MCPs (even if those terms are new to you!)
- Understand and use the core building blocks of Claude Cowork
- Conduct rigorous data analysis you can actually trust — because the AI is working directly with your real data
- Create compelling charts, dashboards, and data visualizations grounded in live organizational information
- Build repeatable workflows and systems that produce consistent, high-quality results
- Automate recurring tasks and multi-step projects so work continues running in the background with minimal oversight
Other benefits
- Learn alongside other nonprofit leaders putting AI to work in their orgs
- Hear use cases from peers that spark ideas you would never have on your own
- Get hands-on coaching from hosts who use these tools every day
Meet your hosts
How we approach AI
- It is a powerful tool to help us get work done
- We only care about AI in relation to what it enables us to do
- We're not AI hype engines
- We're not going to speculate about the future
- We won't try to persuade you to use AI more or less
Why we're starting with Claude Cowork
- Chat, Cowork, and Code are a natural elevation — three options that build on each other
- Most people have chatted with AI but not arrived at "the next level"
- Cowork allows us to store context locally, so we can move our data between tools
We like Claude Cowork today, but we're more interested in applying concepts that are true of all LLMs and will be useful now and into the future regardless of LLM provider.
Cohort overview
Session 1 · July 30
Overview
Working with LLMs requires a new way of thinking. There are infinite use cases but all of them depend on a few foundational building blocks. We’ll cover soft skills like writing and design as well as more technical skills like analyzing data, scheduling automations, and writing scripts. AI can help with all of these building blocks, but human judgment remains the ultimate requirement. We’ll consider how to weave between soft skills, technical skills, and human skills in this session.
Session 2 · August 6
Create your knowledge architecture
When we are intentional about the context we give an LLM, we work with the grain instead of against it. But how do you build this architecture? With care! For AI to “understand” how we work, we must create a map of our work that is easy for LLMs to navigate. This requires 1. Knowing what we do. 2. Being able to articulate it clearly. 3. Storing that knowledge in a logical way.
Session 3 · August 13
Connect your tools
Being able to describe what we do is a great start, but at some point we’ll need to know what we actually did. The work (whether it’s meals prepared, books distributed, or families housed) is almost always measured. You have to count stuff. This data lives in countless databases and spreadsheets scattered across your org. We need to give LLMs access to this data to unlock incredible new skills like data analysis, preparation, and visualization. Everyone is a data analyst now.
Session 4 · August 20
Stack skills together
Now that we have the building blocks established, we’ll begin stacking them together. Most people ask Claude to brainstorm with them or write a first draft. We’ll move from doing one thing to doing multiple things at once. An example might be stacking: Research + Writing + Gathering Data + Visualizing Data to apply for a grant. When you can stack sturdy skills on top of one another, the possibilities compound.
Session 5 · August 27
Where to go from here
We’ve established knowledge architecture and connected our tools, but where do we go from here? People change. Processes change. Models change. We’ll need to adapt. We’ll discuss ways to keep your data fresh, avoid bloated context, and tinker with new processes. We’ll also touch on advanced features like using APIs, writing scripts, and publishing on the internet. We’ll set the foundation for you to experiment in a safe way. We’ll end the session by making an action plan for the next 90 days to integrate these lessons into our work.
Before you join
You should know that you'll need a Claude Pro plan (and willingness to upgrade if necessary). Anthropic offers nonprofit pricing.