Practical AI for Not-for-Profits: the series
Most not-for-profits we work with are about as curious about AI as they are wary of it. They’ve seen the demos. They’ve also read about chatbots giving wrong answers and staff pasting confidential documents into free tools. What they usually don’t have is a clear idea of what a useful first project looks like for an organisation their size, or what it would cost to build and run.
This series aims to give them that. Each post takes one practical use case and works through it the way we would with a client. We cover what the system does and what it needs to connect to. We also cover what it costs to build and keep running, and where it falls short.
Who it’s for
We’re writing for IT managers, operations leads, CEOs and board members at Australian not-for-profits and community organisations. You don’t need a technical background to follow along. We will still include technical detail where it affects cost or risk.
Why NFPs need a different approach
Most AI advice is written for companies that have a dedicated data team and room in the budget for experiments that go nowhere. Not-for-profits face three constraints that change the calculation.
First, they run lean. Staff cover several roles and there’s rarely spare capacity to look after a new system. Anything we recommend has to save more time than it takes to maintain. Licence and usage costs also matter, because they come out of the same budget as service delivery.
Second, they hold sensitive data. Beneficiary records, case notes, health and disability information and donor details all come with legal and ethical obligations, and funding agreements often add more. Before any of that data goes near an AI model, you need to know where it’s processed, whether it’s retained and who can see the outputs.
Third, they answer to boards and funders. A board will reasonably ask what the system does, what happens when it gets something wrong and how it fits the organisation’s risk appetite. Funders may have their own conditions on how data and grant money are used. If a project can’t be explained clearly to both, it will stall no matter how good the technology is. We assess every use case in the series against all three constraints.
Where this comes from
PicNet is a Sydney software and IT services company, established in 2001. We have supported Australian not-for-profits for many years, including Guide Dogs NSW/ACT and a number of other community organisations. Our team is small and senior, and everyone who manages work here is an engineer. We build production AI systems, along with the integrations that connect them to the CRMs, finance systems and case management platforms organisations already use.
That background shapes what you’ll read here. We’ll describe working systems and give real costs where we can. We’ll also say plainly when a use case isn’t worth doing, or when a cheaper fix without AI would do the job better. Sometimes the honest advice for a small organisation is to wait.
PicNet builds production AI systems for Australian organisations, including not-for-profits. If you’d like to work through what a first project could look like for yours, talk to us.