A practical AI starter kit for small healthcare non-profits
Where AI helps a small healthcare non-profit on a tight budget, where it doesn't, and a simple 30-day plan to try it safely without risking client privacy.
Small health and community organizations hear a lot about AI and get very little practical advice that fits their size. You don't need a strategy, a data platform or a consultant to start. You need one tedious task, a few clear rules and a month to see whether it helps.
Start with a chore, not a strategy
Ask your team one question: what do you write, rewrite or summarize every week that you wish you didn't? The answers make the best first projects, because they're low-risk, frequent and easy to measure.
Good places to start
These work well with today's general-purpose AI assistants, as long as no client information goes in:
- Plain-language rewrites. Turn a dense policy, consent form or program description into something your clients can read, at the reading level you choose.
- First drafts. Newsletters, social posts, volunteer emails and event notices. A person still edits and approves, but starts from something instead of a blank page.
- Grant work. Drafting sections from your own program notes, checking a draft against the funder's criteria, and summarizing long guidelines.
- Meeting summaries and action lists, from notes that contain no client details.
- Translation drafts of public materials, always reviewed by a fluent speaker before anything goes out.
- Answering staff questions from your own documents, such as policies and procedures, once you have a tool set up to keep those documents private.
Where not to start
- Anything that involves a client's health or personal information, until you've confirmed the tool's terms, where the data goes, and that you have the agreements your privacy obligations require. The free consumer versions of AI tools are usually not the place for it.
- Clinical or eligibility decisions. AI can help draft or explain, but people decide.
- Anything you can't check. If nobody on the team can tell whether the output is right, it isn't ready to use.
Write a one-page AI policy
It can be short. Cover:
- which tools staff may use, and on which account
- what never goes in: client names, health details, contact information, case notes
- that a person reviews everything before it's sent or published
- who to ask when unsure
A clear page protects your clients and gives cautious staff permission to try.
Keep costs down
Start with the tools you already have. Many office suites now include an AI assistant, and several AI providers offer discounted or donated plans for registered charities. Ask before you buy. A paid business plan is often worth it once you rely on a tool, because those plans usually give you better privacy terms and admin controls than free accounts.
A 30-day plan
- Week 1. Pick one chore and two volunteers to try it. Write the one-page policy.
- Week 2. Use AI for that chore every time it comes up. Keep a simple log: task, time taken, what needed fixing.
- Week 3. Compare the log with how long the task used to take. Note any mistakes the AI made and whether a person caught them.
- Week 4. Decide: keep it, change it or drop it. If you keep it, add one more chore and repeat.
That's the whole kit. A month of honest measurement will tell you more than any vendor demo.
If your organization would like a hand getting started, I volunteer my time for exactly this. You'll find my email at the bottom of the home page.