Problem
An NGO communications team was processing tens of thousands of inbound emails — donor enquiries, partner requests, press, volunteer applications, advocacy correspondence — all flowing into shared inboxes. The team had a small full-time staff who were spending most of their day routing and drafting initial responses to repetitive enquiries.
Approach
LLM-based email management layer. Each inbound message classified across multiple axes — intent (donation, enquiry, partnership, complaint, press, advocacy), urgency, language. Routing to the right team channel with confidence-scored draft responses for repetitive categories (donor receipts, volunteer onboarding) that staff could review-and-send rather than draft from scratch. Multi-language support for an international donor base.
Stack
Claude · Python service layer · email integration · multi-language classification · audit log
Outcome
Communications team shifted from drafting routine responses to reviewing and sending. The team kept full control — the LLM never sent autonomously — but the time saved on routine drafting freed bandwidth for the higher-value advocacy correspondence the NGO actually wanted them on.