An AI assistant that answers questions about a 1,700-member business community in seconds, replacing manual HubSpot lookups.
CUB A private members club for business owners.
CUB's core value is connecting members. But all member data for 1,776 members lived in HubSpot, and every question meant a manual search. Relationship managers were hired partly to do lookups: who should meet whom, who works in construction, whose renewal is coming up, whose birthday is this week. Making one introduction meant digging through profiles one at a time.
Built the chat interface end to end and connected it to the club's HubSpot data, so the team asks in plain English instead of searching by hand.
A relationship manager types "suggest three members Daniel should meet." The AI matches on industry, expertise, and who each member wants to meet, and returns candidates with their member tier. Then "draft an intro email for Daniel and Sarah" produces a ready-to-send draft. Then "who's up for renewal this month" or "whose birthday is this week" returns a list with an outreach reason attached.
Member search and filtering by industry, expertise, revenue and status, an events panel with attendance counts and upcoming events, birthday search, renewal tracking with correct timezones, and member notes with authors. I also built highlight-to-reply: select any name or event in a response and ask a follow-up about it, and the backend fetches that record's data before answering. Plus the interface itself: dark mode, the club's gold branding, mobile support.
Lookups that took minutes of HubSpot digging now take one question. The team makes member introductions and renewal outreach without a manual search step, which is the core of how the club retains members.
All the data here is mocked data. It's for internal use only.