How AI4Love Works

AI4Love is a relationship intelligence platform built on a hub-and-spoke architecture. It doesn't replace your existing systems—it connects them, analyzes the unified data, and surfaces insights your team can act on.

1. Connect Your Data Sources

AI4Love integrates with 23 platforms via OAuth through our Nango integration gateway. Your data stays in your systems—we create a unified view on top.

Donor Management (8) - Blackbaud RE NXT, DonorPerfect, Bloomerang, Virtuous, Classy, EveryAction, Neon CRM, GivingData
CRM & Engagement (5) - Salesforce, HubSpot, NationBuilder, Wild Apricot, Foundant GLM
Email Marketing (1) - Mailchimp
Events & Volunteers (3) - Eventbrite, Galaxy Digital, Better Impact
Finance & Payments (5) - Stripe, PayPal, QuickBooks, Xero, Fluxx
Data Enrichment (1) - Environics Analytics

2. Unified Participation Model

All supporter activity—donations, volunteer shifts, event attendance, engagements—flows into a single Participation table in Airtable. This is the foundation of everything AI4Love does:

People Table - Central identity record (40+ fields including RFM scores and demographic enrichment)
Participation Table - Unified activity stream. Every interaction becomes a participation record.
Insights Table - AI-generated insights written by nightly agents
Computed Fields - Participation scores, activity trends, and velocity metrics calculated automatically by Airtable

3. Nightly AI Agent Analysis

Every night, Make.com orchestrates 7 specialized AI agents that analyze your Participation data:

At-Risk Detection - Flags supporters showing declining engagement
Conversion Opportunities - Finds supporters ready for deeper commitment
Recognition Triggers - Detects milestones worth celebrating
Campaign Intelligence - Surfaces individual-level campaign signals
Relationship Deepening - Finds connection-building opportunities
Cross-Agent Intelligence - Synthesizes patterns across agents
KindMind Enrichment - Adds context from your organization's knowledge base

Each agent uses deterministic pattern detection first, then passes eligible supporters to Claude AI for narrative insight generation. The AI writes the insight; your team decides what to do with it.

4. Four Dashboard Views

Insights are surfaced through four complementary dashboards:

Pulse - Participation analytics with 12-month trend charts across donations, volunteering, and events
Insights - AI analysis visualizations including risk scatter plots, channel effectiveness, and agent distribution
Nurture - Segment health management with five health buckets (Thriving/Steady/Cooling/At-Risk/Lapsed)
Connect - Relationship prioritization with Celebrate, Reconnect, and Appreciate actions plus AI-drafted messages

5. Your Team Takes Action

AI4Love is human-in-the-loop by design. It will never:

Send emails or messages on your behalf
Make calls or contact supporters directly
Modify records in your source systems
Take any action without staff review

Your team reviews AI-generated insights, uses the suggested messages as starting points, and takes action through your existing tools.

Technology Stack

Built on proven, scalable technology:

Frontend - React + Vite with Mantine UI, deployed on Vercel
Backend - Node.js + Express, deployed as Vercel serverless functions
Data Hub - Airtable (multi-tenant, one base per organization)
AI Engine - Claude API (Anthropic) for narrative insight generation
Knowledge Base - Pinecone vector database with OpenAI embeddings
Orchestration - Make.com for nightly agent scheduling
Integrations - Nango OAuth gateway for 23 platforms
Authentication - Authsignal passwordless passkey auth (WebAuthn)

Data Security and Privacy

We take data security seriously:

All data encrypted in transit (HTTPS) and at rest
Multi-tenant isolation—each organization's data is in a separate Airtable base
OAuth credentials owned by each organization, managed through Nango
Your data is never shared with other organizations or third parties
Passwordless authentication via hardware passkeys (no passwords to compromise)