Condition intelligence for sensitive inventory

Know what your inventory is still worth, before it becomes waste.

Cold storage measures temperature, location, and stock count. It never measures the one thing that decides profit, the live condition of each batch. Kryoo turns a two-minute storage walk into a FreshScore and the single action that saves the money.

Live in a paid pilot with Ramada Encore by Wyndham, Gurugram.

2 minStorage walk per day
25 SKUsHigh-waste items scored daily
1 loopPurchase to storage to waste
₹/weekAvoidable waste made visible
Today's FreshScore action list · cold room A
Tuesday · 8:00 AM
28
Paneer, 4 kg received 3 days ago · grade B
Use today or move to prep. ~₹1,900 recoverable.
Use first
41
Spinach, 6 crates received 2 days ago · grade B
Wilting. Cook into today's menu before service.
At risk
63
Tomatoes, 12 kg received today · grade A
Use in normal rotation over the next two days.
Watch
88
Coriander, 3 kg overstocked vs weekly use
Do not reorder this week. Usage is below purchase.
Hold reorder
4 batches flagged ₹3,400 value recoverable today Sent to the chef on WhatsApp
The problem

Value drains silently between deliveries.

A batch can sit inside a cold room, at the right temperature, and still lose its economic life. Nothing flags it, so the loss is found at discard, when the only option left is the bin. Roughly a third of fresh produce value is lost after harvest, and operators run their cold rooms on memory, smell, and manual checking.

Temperature is measured

Loggers prove the room stayed cold. They do not tell you whether a specific batch survived it, or how many usable hours it has left.

Stock is counted

The ERP knows how much is on the shelf. It never knows the live condition of what is on the shelf, or what to use first.

Condition is invisible

The one thing that decides profit lives in one worker's head and leaves when they leave. Kryoo captures it, every day.

How it works

A storage walk becomes the day's decisions.

A worker records a short walk through the cold room on the phone they already carry. Kryoo turns it into batch records, a FreshScore, and one clear action per item. No forms, no new hardware, no change to the workflow.

01

Walk

Two minutes with a phone on the round the worker already makes. Records the batches as they are.

02

Capture

Structured batch records: item, quantity, storage age, location, and visible condition.

03

Score

A FreshScore per batch: remaining usable life and recoverable value, weighed against demand.

04

Act

Use first, prep, move, discount, or hold the reorder. One list, sent to the chef every morning.

The score, honestly

A FreshScore fuses five signals, not one camera.

Condition alone does not decide risk. A batch fading today is fine if it is needed today. Kryoo combines what the worker sees, what the batch endured, what it is, whether it is needed, and what finally happened. Today the read is human-in-the-loop; a vision model automates it as the labelled data grows.

01
Visual condition

Colour, wilt, bruising, mould, leakage, packaging, from the walk.

02
Storage exposure

Time stored, temperature, humidity, handling, and shelf location.

03
Batch identity

Item, supplier, quantity, cost, received date, and expiry.

04
Demand and usage

Menu, occupancy, and prep plan. Risk is condition against need.

05
Outcome

Used, wasted, spoiled, or discounted, the label that trains the model.

What Kryoo measures

Not just freshness. Money leaking out of the storeroom.

Once condition is captured daily, it powers decisions across the whole operation, not just the morning list. Each is a module on the same two-minute walk.

01

Remaining value score

How many rupees are still recoverable from each batch, not just whether it looks fresh.

02

Use-first action list

Use today, move to prep, at risk in 24 hours, overstocked, reorder blocked.

03

Condition-first FEFO

Use the worst-condition batch first, not simply the oldest one on the shelf.

04

Procurement intelligence

Stop over-ordering what spoils. Purchase changes tied to real, measured waste.

05

Supplier scorecards

Which suppliers' goods lose value fastest, with the evidence to renegotiate.

06

Waste and savings report

A weekly rupee view of avoidable waste caught, and where the loss begins.

First proof

A hotel paid us to stop its food from dying.

Ramada Encore by Wyndham, on the Gurugram Dwarka Expressway, runs cold storage and a full kitchen. Summer heat left items unable to hold freshness through routine storage, so they paid for the pilot before we had a product to show. It is a closed loop: purchase, storage, kitchen use, waste, and reorder all in one place, which lets us prove the score end to end.

Paid pilot 14 days 25 high-waste SKUs Human-in-the-loop v0
Pilot · Ramada Encore by Wyndham Gurugram
Live
Scope
One cold room, 25 highest-waste items
Capture
Daily storage walk, scored with the kitchen team
Output
Morning use-first action list on WhatsApp
Baseline
Avoidable waste measured in ₹/week before scoring
Trigger
June and July heat breaking shelf life
First paying customer Closed loop outcome data
Wedge, then ceiling

One walk, every room where value decays.

Food cold storage proves the loop, because spoilage is visual, immediate, and paid for. The same engine reads condition for any goods that lose value in storage. Kryoo owns no cold rooms and no trucks, only the decisions.

01

Hotels & kitchens

Cloud kitchens, caterers, and institutional kitchens with cold storage.

Now
02

Quick commerce & retail

Dark stores and modern-trade fresh, where thin margins make waste fatal.

Next
03

Packhouses & cold stores

Exporters, cold-room operators, and reefer logistics between farm and shelf.

Then
04

Pharma, cosmetics, flowers

Any sensitive inventory that degrades with time, heat, or handling.

Ceiling
Why it compounds

The moat is the outcome data, not the score.

A freshness score is math anyone can attempt. The defensible asset is the labelled record of what actually happened to each batch, condition, action, and result, collected across operators who each see only their own node.

Cross-operator by design

A hotel, a dark store, and a packhouse each see one slice of the same decay problem. A neutral layer learns across all of them.

Labels no incumbent holds

Retail tools see orders and sales. Nobody holds batch outcomes, condition to result, across chains. That record is the training set.

Compounds with every site

Each new site adds walks, labels, and coverage. The score gets sharper and the data gets harder to catch, on the same motion.

See what your cold room is quietly losing.

We run a short baseline audit, then a two-week pilot that scores your highest-waste items daily and reports the avoidable waste in rupees. No new hardware, no disruption.

Emailumesh@kryoo.io
PilotRamada Encore by Wyndham, Gurugram
BasedNew Delhi, India