Applying Tech Org Structures to Nonprofits

I recently learned that Uber is restructuring its org chart to match engineers with subject matter experts, called "agentic pods", to bring AI into every function at Uber.

Here's how it works:
1. Pair engineer and domain experts (marketing, sales, etc.)
2. Understand real workflows
3. Identify high-impact automation opportunities
4. Build agents and workflows against real systems
5. Validate with users, measure impact, and iterate
6. Ship solutions that save time, improve decisions, and scale value

All of that is good and well in the corporate world, but what if you don't have engineering teams?

For small organizations, we can distill this into a simpler model:

1. Elect an AI champion (read: the AI enthusiast who quietly rebuilt a reporting template using AI).
2. Sit down 1:1 with every subject matter expert on staff — programs, development, ops, comms. One question: where does your time actually go?
3. Let the expert drive the conversation; the AI champion doesn't run the workflow, they just know where AI may be able to plug in.
4. Pick one workflow and one metric -- hours spent, turnaround time, whatever "better" would actually mean.
5. Create a small test for that one task.
6. Keep what works, and repeat the conversation in the next department.

For example, translating program outcomes and impact metrics into public-facing language — for donors, annual reports, board decks.

Close that loop using AI to draft the public-facing version from the outcomes data, and programs and development each just validate their half.

Keep it simple and start where you are.