If you sell real estate in India, your website is rarely the first problem. The problem is response speed. A buyer fills a Meta or Google form, then talks to whichever builder, broker, or channel partner calls first. An AI receptionist for real estate should treat every fresh lead like a live conversation, not a row waiting in a spreadsheet.
The real estate workflow that matters most
The first workflow to automate is fresh lead callback. Lead submitted -> AI calls within about 60 seconds -> asks budget, preferred locality, BHK, possession timeline, and site-visit availability -> sends WhatsApp confirmation -> writes the answers to the lead sheet. This is more valuable than a generic chatbot because it reaches the buyer while they are still comparing projects.
Qualification questions to use
Keep the qualification short. Ask: budget range, preferred area, configuration, purpose, timeline, loan status, and site-visit slot. The AI should not interrogate the buyer. It should ask enough to decide whether to book a visit, route to sales, or nurture later. For high-value buyers, the agent should alert a human immediately.
WhatsApp follow-up after the call
After the call, WhatsApp does the memory work: project location, map link, brochure, site-visit time, sales contact, and reminder. If the lead did not answer, the fallback WhatsApp should reference the enquiry politely and offer a one-tap callback or visit slot. Connect this to the speed-to-lead AI calling page and outbound calling feature.
What not to automate
Do not use AI calling for purchased cold lists. Do not let the agent make claims about guaranteed appreciation, legal status, or inventory unless those facts are approved and current. The AI should qualify, answer approved questions, book visits, and escalate complex negotiation to the sales team.
Dvaarik real estate setup
Dvaarik connects one lead source first: Meta Lead Ads, Google lead form, your website, Google Sheet, CRM, or webhook. The AI calls the lead, qualifies the buyer in their language, updates the sheet, sends WhatsApp follow-up, and stores recordings and summaries. Start with one project and one source; expand after the first loop proves conversion.
Frequently asked questions
How much will this cost me per month if I get around 300 Meta leads?
For roughly 300 ad leads a month, a reasonable planning figure is about ₹1,200 in voice credits. Work it out from your own assumption: a qualification call that covers budget, locality, BHK, timeline and visit slot is short by design, so if you assume about 2 minutes a call, 300 leads is roughly 600 minutes, and at ₹2/min with simple whole-minute billing that is around ₹1,200. Add the ₹2,000 per 30 days phone recharge if we supply the number, and the Care Plan at the founding rate of ₹5,000/month — required for the first 3 months, optional from month 4. WhatsApp follow-up messages are billed by Meta directly to your own account with zero markup from us. Full breakdown on the pricing page.
What do I have to pay upfront to get started?
The minimum first credit load is ₹10,000, and setup is ₹40,000 which is currently free for our 10 founding clients. In month one, ₹2,000 goes to the phone number recharge and ₹5,000 to the Care Plan at the founding rate, leaving roughly ₹3,000 for actual calls and chats — about 1,500 qualification minutes. Later top-ups start from ₹500 and credits never expire. The free build is an honest trade: the Care Plan for the first 3 months at the founding rate, plus a real testimonial and short case study once you have results. See the founding client program.
When is an AI receptionist NOT worth it for a real estate business?
If your lead flow is low and you or your closer genuinely answer every enquiry within minutes, this will not move your numbers much — the value of an agent is that it responds instantly when a human cannot, so if nothing is being missed there is little for it to pick up. In that case you are better off spending on ads first and automating once the volume is there. It is also a poor fit if your inventory changes daily and nobody can keep the agent's approved facts current, or if your sales happen almost entirely through walk-ins and referrals rather than paid lead forms. Tell us your actual lead volume and how fast you get to them today and we will say plainly whether it is worth building yet — we would rather tell you to wait than build something that sits idle.
Can the AI call leads for different projects and route them to the right sales person?
Yes — the agent can hold separate scripts, price bands, and locality details per project, and route by whatever the buyer answers. We set it up by tagging the lead source (a specific Meta campaign, Google form, or landing page) so the agent already knows which project it is calling about, then it qualifies and either books the site visit or hands off to the sales person who owns that project. Escalation can be a live transfer during working hours or a WhatsApp alert plus a filled lead row outside them. Start with one project and one source, then add more once the first loop converts — more on the outbound calling feature page.
Is it legal to call leads who filled a form on my ad?
Calling people who submitted their own enquiry to you is a different thing from cold calling a purchased list, and we only do the former. The agent calls your own leads and enquiries, identifies your business at the start of the call, respects calling hours, and stops when someone asks it to. We will not build campaigns over bought databases or scraped numbers — no exceptions, however tempting the list. We cannot give you legal advice on your specific situation, so check with your own advisor for anything beyond this. Our reasoning is laid out in is AI calling legal in India.
My buyers speak Telugu, Hindi and English mixed — will the agent handle that?
Yes, including the mixing — plenty of Indian property conversations move between Hindi, Telugu and English inside a single sentence, and the agent switches with the buyer rather than forcing one language. It covers 22 scheduled Indian languages plus Hinglish, Tenglish and Tanglish, and most world languages too, which matters for NRI buyers calling about the same project from a different timezone. The reason we build it this way is simple: a buyer who is comfortable in their own language is more likely to answer straight on budget and loan status. You can hear it yourself before committing anything — see the multilingual receptionist guide.
The best real estate AI receptionist is a speed-to-lead system. If it does not call the buyer quickly, qualify cleanly, and book the visit, it is not solving the core sales leak.
Send us your current lead sheet columns. We will map the real estate AI calling workflow around them.
Tell us your problemTagged

Written by
Rohith Sriramula
Founder & CEO, Dvaarik AI
A laid-off engineer who went all in on Dvaarik AI — he builds every custom AI voice agent personally. This is written from hands-on work with Indian businesses, not theory.