Agriculture

GPC Group

Meta ads to chatbot to AI calling, learning videos, KYC and auto-invoicing

The client

GPC Group runs Meta lead ads at volume across India and sells agri-inputs business to business, to retailers and distributors.

The challenge

Sales, marketing and accounts each relied on manual effort, and three gaps were hurting qualified pipeline and slowing cash collection.

Sales: leads arrived through a WhatsApp and Instagram chatbot, and a slow or missed human reply let paid leads go cold before anyone called.
Marketing: product content, including farmer education on how to use products, was made by hand and did not reach farmers in their own language.
Accounts: KYC checks and invoicing for new customers and distributors were manual, which slowed onboarding and collections.

What we built

Part 1
Sales
- Meta ads feed a WhatsApp and Instagram chatbot that sends the first reply instantly.
- A missed-reply detector finds leads that go cold without a timely human reply and passes them to an AI calling agent, which dials them directly.
- A follow-up agent retries on a schedule if the first call does not connect.
- Each qualified lead goes to the sales manager, sales associate and sales caller together. Leads still unanswered are re-queued, not dropped.
Part 2
Marketing
- An AI content engine turns the product catalogue into social creatives and copy, then posts them on schedule across GPC's own channels.
- A learning-video generator produces animated, multi-language explainer videos on how products are used, aimed at farmers and retailers.
Part 3
Accounts
- Customer and distributor onboarding runs through an automated KYC flow.
- After verification, each order triggers an invoicing engine that creates the invoice and payment link. Collections sync back to Odoo.

Architecture

Results

At a ₹5 lakh average order value, ₹44 lakh a month in attributed revenue is roughly 9 orders a month. The system surfaces those orders, calls the leads and carries each through to a paid, verified, invoiced account.

Tell us your use case. We will build and run it.