The rescue team had adoption and donation data spread across spreadsheets, intake logs, and PayPal reports.

Animals were staying in care longer than expected, and fundraising felt reactive but they couldn't see why.

Volunteers spent hours reconciling records instead of caring for animals.

Donor outreach relied on guesswork. Success stories sparked one-time generosity, but most donors never returned.

I consolidated adoption, intake, and donation records into a working dataset.

Then I mapped intake dates to adoption timelines, length of stay by age group, and donation timing relative to success stories.

Patterns emerged:

• older pets stayed nearly twice as long,

• and donors gave right after seeing stories but rarely came back.

AI entered as a quiet partner.

A lightweight model rewrote long-stay bios using story-first language and suggested photo approaches for overlooked pets.

Clustering analysis identified which messages drove repeat giving.

The system surfaced what humans couldn't see fast enough.

• Adoption bios were standardized using proven story frameworks.

• Donor follow-up emails were automated based on engagement behavior.

• Fundraising posts were scheduled during historically high-response windows.

• Volunteer workflows no longer depended on spreadsheets or memory.

The system handled the data. The humans focused on the animals.

• 26% increase in adoptions

• 12+ volunteer hours reclaimed per week

• Recurring donors doubled

• $18,500 in new donation revenue over four months

• Adoption timelines shortened by nearly a week!

The data created confidence to stop guessing, confidence to automate responsibly, and confidence to focus on what mattered most which then Helped more animals find their forever homes.