Client project — voice AI agent for commercial real estate lead qualification
Who this is for
Commercial real estate brokerages that get inbound calls from buyers, sellers, and investors — where every call needs the same qualifying questions asked before a human agent should get involved, and missed calls mean missed deals.
The problem
Qualifying a real estate lead over the phone means asking the same structured questions every time — name, contact info, buyer or seller, asset class, budget — then manually logging it, checking calendar availability, and booking a follow-up. Doing this consistently for every inbound call requires someone always available and always thorough.
What I built
A voice AI agent (built on Vapi) that handles the full call-to-booking flow:
Tech stack
Vapi, Anthropic Claude (conversation model), Deepgram (speech-to-text), Google Calendar API, Supabase, HubSpot CRM
The result
.jpeg)