Fill a food bank's volunteer shifts in seconds. The AI agent builds an auditable match plan, drafts warm confirmation messages, and writes help-needed broadcasts for any gap it can't fill.
Food banks and small nonprofits chronically have unfilled volunteer shifts — not because volunteers don't exist, but because manually matching volunteer availability and skills to open shifts (via spreadsheets or group texts) is slow and error-prone for a coordinator who is already stretched thin. This agent does that matching and outreach for them.
A deterministic Python @tool assigns volunteers to shifts by role + availability — reproducible, not an LLM guess.
The LLM drafts a friendly, ready-to-send confirmation message for every assigned volunteer.
For any shift it can't fully fill, it writes a group-text-ready "we still need…" broadcast describing the exact gap.
Outputs both machine-readable JSON and a clean, colour-coded visual summary.
Volunteer data + open shift data go into the agent.
The agent calls the deterministic match_shifts tool for the assignments and gaps.
Amazon Bedrock drafts the confirmation and help-needed messages.
You get the plan, confirmations, and broadcasts — as JSON and a visual summary.