Initializing portfolio

000

Aravind.
All projects

Case study 05 / 16

AI Concierge for Lead Capture

Built a conversational AI concierge that engages website visitors, qualifies buyer intent, and captures leads around the clock. The agent schedules viewings and routes hot leads directly to agents, cutting response time and lifting conversion rates.

Case study
05 / 16
Sector
Real Estate
Stack
4 technologies
Published
2026
AI Concierge for Lead Capture

The problem

Property enquiries arrive at all hours and go cold fast. A visitor browsing listings at eleven at night is a real lead; by the time an agent replies the next morning they have often contacted three other agencies. Staffing that window is expensive and mostly idle.

How it works

The concierge handles the conversation, not just the capture. Built on LangChain, it engages visitors, asks the qualifying questions an agent would — budget, area, timeline, intent — and books viewings directly rather than promising a callback.

Twilio carries the conversation across channels so a lead is not lost because someone preferred to text. Qualified leads land in HubSpot with the conversation attached, which means the agent picking it up sees what was already discussed instead of restarting from scratch.

What shaped it

Qualification is the point. An agent inbox full of unqualified enquiries is not obviously better than an empty one, so the design prioritised routing genuinely hot leads immediately over maximising captured contacts.

The agent also had to know its limits — a visitor asking something it should not answer gets a human, with context, rather than a confident guess about a property.

Outcome

Response time to new enquiries dropped sharply and conversion improved, with agents spending their time on leads that had already been qualified.

Let's keep in touch

Get in touch