Who this is for
AI training for NGOs and social service organisations
For social workers, service leads and programme staff. Where headcount is thinnest, what to automate first.
Funding applications and outcome reports, case notes, and community publicity are common and extremely labour-intensive in social service work — and labour is usually the scarcest thing available. This session covers the practical use of AI on those three, the handling rules for service users’ personal data, and which task to automate first when resources are limited: not the most irritating one, the most regular one.
What the session covers
- Draft funding applications and outcome reports from service data
- Organise case notes, and where the line sits on service-user data
- Community publicity copy and a weekly content calendar
- A priority rule for what can be handed to AI and what needs human review
Track record
July 2026 — delivered the "AI+" taster for schools and social service organisations at HKCT; June 2026 — taught an AI class for grassroots women for a Greater Bay Area family foundation.
Common questions
Can service users’ personal data be processed with AI?
Names, addresses, ID numbers, contact details and identifiable case details are the red tier and should not enter any tool the organisation has not provided. The workable path is to de-identify first — rewrite the case as an anonymous scenario.
With limited resources, what should be automated first?
Judge by regularity, not by irritation. A monthly outcome report in a fixed format is a better first target than casework that differs every time: a regular task is set up once and reused, an irregular one needs fresh judgement each round.