01
Ask: "Why did Mumbai stores drop 12% WoW?"
Brain returns the cohort, channel and campaign drivers in seconds.
The AI brain that powers segmentation, personalisation, churn prediction, send-time optimisation and natural-language data queries across the Fundle stack.
Section 01
The thesis
The problem
Every modern vendor claims "AI" — usually a bolt-on feature on top of a 2010s platform. Real AI in loyalty means models running in production, learning continuously, and making decisions in milliseconds — for every member, every transaction, every campaign.
The Fundle approach
Fundle Brain is the AI layer underneath every product. Ten autonomous agents — Loyalty, Engagement, Analytics, Campaign, Retention, Segmentation, Monetisation, Personalisation, ROI Tracking, Store Support — coordinated by a central orchestration agent and exposed through a natural-language interface.
Section 02
The capabilities
10 production AI agents (see full list)
Continuous learning from every interaction
Vector-store memory of member behaviour
LLM-powered natural-language data interface
Model registry with versioned A/B rollouts
Explainability: every recommendation has a "why"
Drift monitoring & auto-retraining
In production
Use cases
01
Brain returns the cohort, channel and campaign drivers in seconds.
02
AI suggests audience, channel, send-time and creative; team approves with one click.
Questions
Most asked
Best-of-breed by task: Claude Sonnet for content/reasoning, GPT-4-class models for tools/agents, in-house embedding model for member memory. Customers can BYO models for enterprise.
No raw PII. All LLM calls use de-identified context. Enterprise plans support self-hosted models within VPC.
Continue exploring
Next step
A 30-minute working session with a Fundle loyalty strategist and a solutions engineer — tailored to your brand or mall.
Hi 👋 I'm Abhinav
Got a loyalty or ADSR question?