Artificial intelligence in nonprofit organizations isn’t about sci‑fi solutions, but about small improvements that free up the hands of people on the front line. Radka Bystřická from the organization Digivia (formerly Sdružení Via) showed how these tools save time, rein in bureaucracy, and improve services for clients. The examples range from writing emails and technical support to care planning and fundraising.
Common yet useful: from emails to meeting minutes
Most often it’s the absolute basics: informational emails, newsletters, or event invitations. If AI gets a quality brief – who you are, whom you’re addressing, and what action you want people to take – it can produce text that would take an hour by hand in ten minutes. The trick isn’t “hacks,” but giving the tool good input data. It’s also important to use versions of the tools suitable for working with sensitive data.
A practical tip is technical support: upload a screenshot with an error to the AI and get advice on the fix instead of waiting for the helpdesk. In office routines, transcripts and summaries of meetings work great as well, from which you can directly create proposals or task logs. According to the trainer’s experience, this saves at least four hours a week, roughly a tenth of working time. This also changes the habit: everything important is transcribed and then the outputs are processed in AI.
Stories from the field: social services, grants, and campaigns
In a hospice, with the help of two graduates of the „AI manažérka“ course from Čekita, a system was created that automates the processing of visit records directly in the client record system. The project reports savings of around 30 % of social workers’ time, realistically at least a solid quarter per day. The organization Neposeda, together with Charles University, created a procedure that also generates secondary documents from the records, especially individualized care plans. In addition to a 10–15 % time saving, this brought higher‑quality, tailor‑made documents and less need to “copy and paste.”
AI also helps with grant applications: it won’t invent a project, but it will help formulate it precisely and clearly, often in a foreign language as well. At Digivia, the author used it to write a high‑quality application for a global call and succeeded also thanks to the right “grant language.” In a recent public campaign, they used ChatGPT (the Pro version) for sociological research, creating segments and personas, subsequent testing of claims and texts, and rapid production of materials; the whole round took approximately 6–8 hours of net time over three days. In general, after half a year of sensible digitization and the adoption of AI, nonprofits often achieve a 10–20 % time saving, about a 10 % increase in client visits, and fewer overtime hours.
