An omnichannel WhatsApp + in-app assistant that knows every member by name — and when to escalate.

A multilingual conversational AI for a senior-citizen membership club, delivering tailored answers across WhatsApp and in-app, with memory, tone-matching, and clean escalation to humans for finance and medical questions.

92%of first-touch queries resolved by AI

4.8/5member CSAT, up from 3.9

<45smedian response time, 24/7

The challenge

What we were up against

GoHappy's members range from 55 to 92, message in 6 different languages, and ask everything from event RSVPs to insurance claim status. The support team was overwhelmed and response times were creeping past 8 hours.

Our approach

How we got from zero to live

01Built a unified brain over the club's CRM, event calendar, knowledge base, and member profiles.

02Per-member long-term memory so the bot remembered last week's questions, preferences, and family details.

03Tone-detection layer that mirrored each member's pace, language, and formality.

04Strict escalation rules: any finance, insurance, or medical query routes to a named human within 2 minutes.

The solution

What we shipped

Live across WhatsApp Business and the in-app inbox with a shared conversation history.

Replies in Hindi, English, Marathi, Tamil, Bengali, and Telugu with code-switching support.

Voice-note transcription so members can speak instead of type.

Weekly retraining loop using flagged conversations and CSAT signals.

Tech stack

WhatsApp Cloud APIClaudePineconeSupabaseLangChainWhisper

PreviousSouthwest Gases

NextMulti-tenant

Let's build your AI advantage

Ready to put AI on payroll?

Book a 30-minute discovery call. We'll walk through your workflows, surface the top 3 opportunities, and tell you whether we're the right fit — no slides, no sales theatre.

Book a discovery call Open services booklet

loading_agent…