The whole operation
What actually happens between your message and your gate
No part of this is clever. It is just ordinary logistics run in the right order — buy after the order, not before it — which is exactly why the produce comes out fresher and the cost comes out lower.
How it works
Four steps, and you only do two of them.
The whole thing runs inside a WhatsApp chat you already have open. Order by 11:00 PM, collect between 7:00 AM and 9:00 AM — the two steps in between are ours.
- 01
Tell us tonight
Scan the QR at your gate or tap Order on WhatsApp. Send a list, a voice note, or just "same as last Sunday." English, Hindi or Telugu — all fine.
- 02
Confirm by 11:00 PM
We price your list, send it back itemised, and you confirm. Pay by UPI on the link in the same chat. No app, no account, no card saved.
- 03
We buy at dawn
At 4 AM we are at the mandi, buying against confirmed orders only. Your vegetables are picked out that morning — not pulled off a shelf they have been sitting on for days.
- 04
Collect at 7–9 AM
Your basket arrives at your society drop point, labelled with your block and flat, roughly three hours after it left the mandi. We take the empty basket back on the next run.
Behind the scenes
Your night, our night
A single day at Rozbasket, from the first order to the empty basket coming back.
- 6:00 PM – 11:00 PM
Orders come in
Residents send their lists on WhatsApp. Our ordering agent reads plain English, Hindi or Telugu — typed or voice — prices it, and sends back an itemised confirmation. Nothing is charged until you approve it.
- 11:00 PM
Cut-off. Demand is now known.
Every order for tomorrow is locked. At this moment we know exactly how many litres of milk, kilos of tomatoes and dozens of eggs each society needs — not an estimate, an actual number. This is the single fact the entire cost structure depends on.
- 11:15 PM
Procurement plan generated
Confirmed demand is aggregated across every society into one purchase list, matched against supplier prices and minimum-order quantities, and split into the cheapest valid allocation. A small safety buffer is added per item based on how often that item's demand has historically overshot.
- 4:00 AM – 5:30 AM
Buying at the mandi
We buy against the plan. Perishables are graded on arrival — presentation-grade produce goes to households and restaurants that asked for it, second-grade goes into discounted baskets rather than the bin.
- 5:30 AM – 6:15 AM
Packing by society, block and flat
Orders are packed into reusable branded baskets, grouped by society, then by block, then by flat number. The route sheet and the packing order are the same order, so nothing is searched for in the dark.
- 7–9 AM
One vehicle, one society, one stop
The vehicle reaches your society gate, drop point or clubhouse. Baskets come out in block order. A society of 80 orders is unloaded in minutes, and empty baskets from yesterday go back with the vehicle.
- 9:30 AM
The loop closes
Anything unsold is routed to a buy-back arrangement or a local partner instead of being written off. Every order, substitution, complaint and leftover kilo feeds tomorrow night's forecast.
The economics
Why density beats speed
The metrics we optimise are not delivery time. They are hours from mandi to gate and cost per delivered order. Every design decision on this page pushes one of them down — and the same decisions push both, because buying only what is already sold is simultaneously the freshest way and the cheapest way to run a perishables business.
- Shared trips. One vehicle serves an entire society. The delivery cost of a single order is a fraction of a trip, not a whole one.
- Zero-guess buying. Perishables are bought after demand is known, so spoilage is a rounding error rather than a line item.
- One drop point. No lift-by-lift door delivery unless a society asks for it. Time per society falls, orders per route rise.
- No dark stores. Nothing sits on a shelf waiting to be wanted. Stock is bought, packed and delivered inside one morning.
- Recurring orders. A standing weekly basket is demand we know a week early, at effectively zero acquisition cost.
The forecasting loop
Once a society has ordered for a few weeks, its consumption pattern is remarkably stable. We forecast per society, per item, per day using order history, day of week, weather, festivals, salary dates and seasonality.
Indicative weighting of the signals our procurement model uses. Accuracy improves as a society builds order history.
More on how we build this →Order tonight. Fresh at your gate by 7 AM.
Bought at the mandi at dawn, never off a shelf. No app, no minimum order, no delivery fee — just message us before 11:00 PM.
Or call us on +91 76758 54788