Sondhan · সন্ধান · the inquiry

Find the fraud in your distribution before it reaches the field.

Counterfeit and diverted product costs agrochem manufacturers margin, brand, and farmer trust — and unit-level authentication is too expensive for a 50kg bag and unenforceable at a rural counter. Sondhan takes a different route: instead of trying to authenticate every unit, it watches how your product actually moves through distribution and surfaces the anomalies fraud leaves behind — so your regional teams investigate on evidence instead of rumor.

Built for fertilizer, pesticide, and seed manufacturers across India and Bangladesh.

The problem
Counterfeit and diverted agrochem erodes margin and reputation — and a farmer who buys a fake input blames your brand when the crop fails.
Why auth fails
Unit-level tags cost more than the bag and depend on rural retailers scanning every unit — which doesn't happen at scale.
The approach
Watch the movement of product, not the authenticity of each unit. Fraud distorts normal distribution in ways that stand out — once you're looking at the whole channel.
The core idea

A scan that doesn't match is not proof of a fake.

Counterfeit operators don't print working QR codes, and a genuine bag whose scan simply never happened looks identical to a fake one. The signal isn't in any one scan — it's in the shape of the whole channel.

Unit authentication

Verify every bag — in theory

  • A cryptographic tag per unit costs more than the product it protects.
  • Depends on every rural retailer scanning every unit, every time.
  • A missing scan is ambiguous — real product unscanned looks just like a fake.
  • Counterfeiters simply don't participate, so the gap tells you nothing.
Pattern detection

Watch how product moves

  • Uses the distribution records you already generate — no exotic hardware.
  • Fraud shows up as movement that doesn't add up across your channel.
  • Those anomalies become ranked, investigable cases.
  • Your regional team confirms on the ground — the system points, people decide.
What it looks for

Fraud can't move product without leaving a trace.

Sondhan watches for the handful of ways fraud bends a normal distribution pattern — product surfacing where it was never shipped, moving faster than physical delivery allows, or concentrated in places the numbers don't support. Each anomaly it finds becomes a ranked case your team can act on. The specific signals, and the thresholds that trigger them, are tuned to your channel.

What Sondhan is — and isn't

The system points. People decide.

Sondhan is an intelligence layer, not an enforcement engine. It will not tell you a bag is counterfeit, and it won't act on your behalf. What it does is turn millions of commission and scan events into a short list of ranked anomaly cases, each one backed by the specific numbers that fired it.

From there, the investigation is human: your regional sales team makes the call or the field visit. Sondhan's job is to make sure that when they do, they're spending their time on the cases that actually warrant it.

How it works

From handoffs and scans to a case worth a phone call.

You already generate most of the data. Sondhan captures it, watches it, and surfaces what's worth a human's attention.

1

Capture how product moves

Each unit carries a cheap serial code. Sondhan records how product moves — factory to distributor to retailer — and the checks that happen at the counter.

2

Watch for anomalies

Sondhan continuously compares live movement against what's normal for your channel, using thresholds your operations team controls.

3

Open ranked anomaly cases

When a pattern crosses its threshold, Sondhan opens a case carrying the evidence — which container or retailer, which rule fired, the numbers behind it.

4

Hand off to the field

Cases route to the right regional team for ground-truth investigation. They confirm, dismiss, or escalate — and that outcome sharpens what the system learns to surface next.

Who it's for

Built for manufacturers who own their whole channel.

Sondhan is sold to the manufacturer and runs entirely inside your own tenant — there's no shared graph and no upstream data to pull. The chain is yours end to end, and so is the intelligence Sondhan draws from it.

Fertilizer makers

High-volume, low-unit-cost product moving through long rural distribution chains — where diversion and substitution thrive.

Pesticide & crop-protection

Premium branded inputs counterfeiters love to imitate, where a fake product damages crops and the brand takes the blame.

Seed companies

Branded seed packs sold across fragmented retail, where channel integrity directly determines what farmers actually plant.

See the fraud your distribution data is already showing you.

Tell us how your product moves from factory to field, and we'll walk through what Sondhan surfaces across your channel.

Sondhan detects statistical anomalies in distribution data to guide human investigation. It does not authenticate individual units and does not by itself establish that any product is counterfeit; findings are leads for ground-truth investigation, not determinations.