IndiaMART’s Agentic AI Now Handles 1 Lakh Buyer Calls a Day
IndiaMART has built one of India’s biggest live business AI systems. Its agentic AI voice tool, called IM VANI, now handles close to 1 lakh (100,000) buyer-seller calls every single day. These are real calls that connect buyers with sellers on India’s largest B2B marketplace. B2B means “business to business” — companies selling to other companies. The system grew from just 5,000 to 10,000 calls a day to this massive scale. It shows how far AI voice agents have come in India.
So what is “agentic AI”? It is AI that can take actions and complete tasks on its own, not just answer one question. Here, the AI makes calls, talks to buyers, understands their needs, and moves the deal forward. IndiaMART built it with telecalling partner SquadStack.ai. Together they call it India’s largest agentic AI system in live commerce.
What is IM VANI and how does it work?
IM VANI is a voice AI that speaks with buyers to confirm what they want. When a buyer shows interest, the AI calls them. It checks the requirement, then helps match them to the right seller. This is the core job of IndiaMART: good buyer-seller matchmaking. Amarinder S Dhaliwal, Chief Product Officer at IndiaMART, said the first priority always stays accurate matchmaking.
The AI is trained for India’s messy, multilingual reality. It handles Hindi, English, “Hinglish” (a mix of Hindi and English), Tamil, and other regional ways of speaking. It is also trained on nearly 1 lakh B2B product categories. That breadth lets it understand a huge range of buyer requests.
The results: faster, cheaper, and surprisingly human
The numbers are striking. IM VANI automates about 95% of buyer voice conversations. It delivers 20% higher conversion than manual calls. Conversion means turning an interested buyer into a real lead or sale. The AI also connects with 75% of leads, compared with 50% for human agents on the same leads. And it cut the cost per confirmed lead by 15%. Overall, customer acquisition cost dropped by 2 to 3 times versus human-only operations.
Most striking of all, people often cannot tell it is a machine. At the Global Fintech Fest 2025, a blind listening test was run with 1,563 BFSI leaders. BFSI means “banking, financial services and insurance.” In that test, 81% of them thought the AI conversations were human.
Key facts
| Metric | Figure |
|---|---|
| Daily calls handled | Nearly 1 lakh (100,000) |
| Starting volume | 5,000–10,000 calls a day |
| Automation rate | 95% of buyer voice conversations |
| Conversion vs manual | 20% higher |
| AI connectivity rate | 75% (vs 50% for humans) |
| Cost per confirmed lead | 15% lower |
| Customer acquisition cost | Down 2–3x vs humans |
| B2B categories trained | Nearly 1 lakh |
| “Sounded human” test | 81% of 1,563 BFSI leaders fooled |
What happens to the human workers?
The AI does not do everything alone. About 80% of use cases are handled fully by AI. The other 20% are harder cases. Of those, roughly 30% are passed to human agents. So humans still matter — their role just changes. They move from making calls to overseeing the AI. They check quality, audit calls, and handle tricky cases. IndiaMART calls this a shift from “manpower-scaled” to “intelligence-scaled” operations.
Safety guards are built in. The system uses confidence thresholds, post-call audits, and WhatsApp follow-ups. A confidence threshold means the AI only acts when it is sure enough; if not, a human steps in. Apurv Agarwal, CEO and co-founder of SquadStack.ai, said the product was shaped entirely by what IndiaMART’s buyers actually need.
FAQ
What is agentic AI in simple terms?
Agentic AI is AI that can act on its own to finish a task. Instead of just replying once, it can make calls, ask questions, and move a job forward step by step. IndiaMART’s IM VANI is a working example.
Will this AI take human jobs at IndiaMART?
Roles are changing, not just vanishing. Staff move from making calls to supervising the AI, auditing quality, and handling complex cases. The hard, human-judgment work stays with people.
Can the AI speak Indian languages?
Yes. It handles Hindi, English, Hinglish, Tamil, and other regional speech. This helps it talk naturally with buyers across India.
Why it matters (especially for India and founders)
This is a landmark moment for Indian AI. It proves voice agents can work at huge scale in India’s many languages — and pay for themselves. Cutting acquisition cost by 2 to 3 times is a game-changer for any business. For founders, IndiaMART shows the path: pick a real, repetitive task, partner with the right AI firm, and scale carefully with guardrails.
It is part of a clear wave of practical AI across India. We see the same shift in finance, with a startup turning financial advisors into AI-powered wealth coaches, and in big global firms, like Bristol Myers Squibb scaling AI and data analytics from Mumbai. Indian companies are not just testing AI. They are running it in production at scale.
The takeaway
IndiaMART’s IM VANI shows agentic AI is ready for prime time in India. It handles 1 lakh calls a day, lifts conversion, and slashes cost — while sounding human. The lesson for every Indian business is clear. Voice AI is no longer a demo. It is becoming a core engine for sales and service at scale.
