કોઈ ભૂખ્યું નહીં સૂવે · Reverse logistics for surplus
We built rails that move surplus goods backwards. The first cargo is the hardest one.
A restaurant, a temple kitchen, a festival pandal or a household with leftover vegetarian food taps once. Our dispatch system sends riders out with empty containers, back with filled ones to an FSSAI-licensed centre, and out again with hot thalis to shelters and people who need them. Free at both ends, funded by advertising. That is AnnSetu, live in Ahmedabad since 4 September. The engine underneath it is not specific to food.
Live · 3 centres · Ahmedabad · 100% automated dispatch
Live on the apps. Live on the streets.
This is a software company with a logistics operation attached, not the other way round.
A consumer app on iOS and Android, a separate rider app, an admin dashboard, and a dispatch
engine that composes three legs from one inbound order — running on our own edge-deployed
backend with four carrier integrations behind it. That is why the 200-meal collection on
25 September needed no one to coordinate it.
Both live. Rider app on Play as well.
The number that matters most
5%
Our three licensed centres are running at roughly five per cent of capacity. The buildings, the licences, the cold chain and the staff are already paid for. Growing twenty-fold requires signatures, not capital expenditure — and every additional meal is close to marginal cost.
₹2.97
Cost to collect one meal, at 200-meal scale
₹10.65
All-in variable cost per meal delivered
200+
Meals from one automated collection, 25 Sep
25 min
From a partner's tap to a rider at their door
30+
Own riders, plus 3 national carriers on automatic failover
<5%
Discard rate. Everything else is eaten, same day
Rider cost per meal has fallen from ₹29.90 on our first 49 meals to ₹5.47 — while we added a third leg to the operation. Collection cost is driven by distance, not volume, so it barely moves as batches grow. Repeat partners on the supply side, repeat recipients on the demand side, and word of mouth we have not paid for.
The mechanism
LEG 1 · REVERSE
Containers out
A rider carries empty container kits from the centre to the supply partner. This leg is why large pickups are possible at all — a restaurant closing for the night, a festival pandal or a temple kitchen sitting on 200 meals does not have 200 containers. Household partners pack into their own boxes; the kit is what unlocks everything above a few kilos.
LEG 2
Goods back
Filled, sealed, batch-labelled containers return to a licensed centre. Validated cold chain from the moment of pickup.
LEG 3
Delivery out
Sorted, portioned and dispatched to recipients on batched runs, with a delivery record against every unit.
All three legs are composed automatically from a single inbound order. Human involvement is limited to centre staff. No one-way delivery platform can do this — they have no centre, no container asset and no reverse leg. That outbound-empties capability is the whole company.
25 minutes
Partner taps → rider at their door
40 minutes
Partner → licensed centre, cold chain from pickup
15 minutes
Centre → rider collected for delivery
35–40 minutes
Centre → the person who eats it
The national layer is already built
4 carriers
Our own riders, with Delhivery, Shiprocket Quick and Porter behind them. AnnSetu riders serve all of Ahmedabad; when none is free the backend routes the ride to a national carrier automatically, white-labelled, so the supply partner and the recipient see an AnnSetu rider throughout. The network cannot run out of capacity, we carry no idle fleet, and we are not dependent on any single carrier.
This is a routing layer, not a fleet. Three consequences. We do not decline an order for want of a rider — the service level a sponsor is buying, and the one thing a volunteer network structurally cannot offer. No carrier has pricing power over us, because the system routes to whoever is available and cheapest. And movement in the next three cities is already solved and tested: Shiprocket Quick performs strongly in the metros even though it is weak here in Ahmedabad, which is precisely the coverage profile city two needs.
So opening Mumbai means an FSSAI-licensed centre, local supply, and a recipient network — three things that took us weeks here without a playbook, and which we now have a playbook for. Everything around them is already automated. Ahmedabad to 1,000 meals a day, then Mumbai, Bangalore and Delhi.
Who pays, and why that is the difference
Every comparable to date
Charge the supplier
US food-rescue companies bill the business giving the goods away, paid out of a tax arbitrage: under the PATH Act a donation writes off roughly $200 against about $100 for disposal. It is discretionary, capped, and has produced a hard ceiling — the two closest comparables raised about $4M and $8M across a combined eighteen years and have not raised since.
Payer: the supply side
AnnSetu
Neither side pays
Goods move free in both directions. Revenue comes from brands buying association and attention, and from the people who want the work to continue. That is a marketing budget — larger than philanthropy, renewable annually, and bought on reach we can measure and prove was delivered.
Payer: a third party, for advertising
We are an advertising-funded utility with a logistics engine underneath — the structure of free radio and free search, applied to surplus. A brand's name on a thali handed to someone at their lowest moment is a different asset from a billboard, and it arrives with a delivery receipt for every impression.
Revenue
Naming rights
₹1 lakh per month per centre; ₹4 lakh per month for a city title, on three-to-five year terms with escalators and right of first refusal.
In-app advertising
Placement inside the consumer and rider apps, against an audience that opens them because they are doing something they feel good about.
AnnSetu Ads
₹599 per day, self-serve. A neighbourhood business appears on the packaging and in-app across the meals it funds that day.
Setu Moments
The impact report against meals paid for — where they went, when, and to whom — issued across every city we run in. It is what converts a sponsorship into a renewal.
Subscriptions
₹30 per month, live and growing. Meals are always free to whoever receives them. “Hungry won't pay. Satisfied will.”
Societies
Residential society partnerships billed as a line on the maintenance bill. The loop is already running by itself: residents subscribe, and the security guard at their own gate eats. The person paying ₹30 a month and the person eating the meal are in the same building.
Why the rails matter beyond food
India now regulates eight waste streams under Extended Producer Responsibility, with 4,574 registered recyclers processing 41.76 million tonnes as of March 2026. Textile EPR draft rules are expected within FY2026, and the Textile Ministry's own March 2026 waste-mapping report flags gaps in collection infrastructure. Goodwill spending is discretionary and can be cut in any quarter. EPR creates a statutory obligation to collect — and collection is precisely what our reverse leg does.
AnnSetu
Live
Surplus food. The hardest cargo: perishable, safety-critical, hour-bound, cold chain.
GyanSetu
Built · held
Books, circulating and returned. Ships when AnnSetu is sustained at 300 meals a day.
VastraSetu
On the EPR clock
Clothing and factory offcuts, timed to the textile EPR rules.
ToySetu
Later
Toys, on the same collection and redistribution rails.
These are sequenced, not simultaneous. Each one is deliberately dormant until the one before it holds a stated number. If we can move hot food safely within the hour, everything else on these rails is an easier problem. India's reverse-logistics market was $50 billion in FY2025, and no one is serving the surplus-redistribution end of it.
Where we are
Operating
Three FSSAI-licensed centres covering most of Ahmedabad, at ~5% utilisation. Repeat partners on supply, repeat recipients on demand, organic word of mouth on both.
Supply, organic
Four channels are already sending, none of them prospected: restaurants at close of service, festival pandals — we collected from Ganesh Chaturthi — temple kitchens, and households packing their own boxes. Every one of these came to us. The festival and temple calendar makes a share of this supply recurring and predictable in a way no Western comparable has — Navratri and Diwali are both ahead of us this quarter.
Technology
Consumer app live on both the App Store and Google Play, a separate rider app, an admin dashboard, and an automated multi-leg dispatch engine — all in production, on our own edge-deployed backend, with Delhivery and Shiprocket Quick integrated and Porter in progress behind an automatic routing layer. Built and shipped by a single founder before the first outside rupee. The 200-meal collection on 25 September ran end to end with no manual coordination.
Safety
Vegetarian only. Serving-dish sourcing, never plate returns. Validated cold chain with temperature logs. Discard on doubt, to biofuel.
Demand
We almost never hold stock. Everything goes out, the same day, with under 5% discard. Five community shelters tagged and receiving daily, more in the pipeline — and a stream of individual orders from security guards across residential societies, eating every day.
Revenue, unsold
Monthly and yearly subscribers, and a first paying Setu Moments customer buying an impact report against meals they funded. Neither product has ever been pitched. The numbers are small; the sales effort behind them was zero.
Constraint
Supply. Not demand, not infrastructure, not unit economics — we place every meal we collect. The container loop removes the reason a partner with real volume could not say yes; turning organic inbound supply into signed, scheduled repeat partners is what the raise buys.
AhmedabadFrom 200 to 1,000 meals a day — a partnerships team to fill capacity we already own and pay for.
Three citiesLicensed centres in Mumbai, Bangalore and Delhi. Movement is already solved; this buys premises, licences, local supply and each city’s cold start.
ProofTwelve months of audited impact and unit-cost data across four cities — what prices the Series A, the EPR contracts and the naming rights.
$750K funds Ahmedabad to 1,000 a day plus two cities. $1.25M funds all three cities with eighteen months past the third opening. The difference is each city’s cold start: collection cost begins at ₹29.90 a meal and falls to ₹5.47 only once density is paid for, and Ahmedabad is the only place that curve has already been walked. At the end of this capital: four cities live, Ahmedabad at 1,000 meals a day at or below ₹20 all-in, a signed city-title sponsor and a real subscriber base. Structured as a convertible so the price is set by what the cities actually do, not by what we project today.
AnnSetu is operated by The Setu Ecosystem (OPC) Private Limited · CIN U56290GJ2026OPC182611 · Incorporated 19 August 2026. Operating figures are measured from live operations between 4 and 25 September 2026 and are unaudited.