Food Service
Reduction in perishable wastage
Fill rate achieved, up from below-target levels
About
The operator aimed to protect its chilled supply chain while tightening procurement. A 3% daily shrink rate on perishables and nightly email reports left managers reacting a day late, causing emergency buys and lost margin. We built Kafka streams feeding a live freshness dashboard, with auto-replenishment triggered the moment stock ran low.
Industry
Food Service
Company size
1,000 – 5,000 employees
Founded
1996
The Company
A chilled-chain operation reporting a day behind reality
The company operates a national food-service supply chain built around chilled and perishable inventory, where shelf life is measured in days and a stock decision made too late is a stock decision that becomes waste.
Inventory visibility ran on nightly email reports summarising the previous day's positions — useful for a retrospective view, but a full cycle behind the real-time decisions managers actually needed to make.
The challenge
Reacting a day late to a chain that moves in hours
A 3% daily shrink rate on perishables compounded across the network into a material margin drag, and because managers only saw yesterday's positions, they were consistently reacting to a shortage after it had already forced an emergency buy at a worse price.
Procurement decisions were similarly lagged — reorder triggers were manual and dependent on someone noticing a low-stock report in time to act, rather than the system flagging it the moment a threshold was crossed.
The Solution
A live freshness clock and automatic replenishment
We built Kafka streams feeding live Tableau dashboards showing a freshness clock for every SKU, replacing the nightly email report with a continuously updating view of shelf life and stock position. When a Holt-Winters demand model signals a SKU is running low, an API automatically pushes a replenishment order into Microsoft Dynamics NAV.
Managers went from reading a report about yesterday to watching a live clock counting down on every perishable item in the network, with the reorder decision already triggered before they even needed to look.
The Results
18% less wastage, 96% fill rate
Perishable wastage fell 18% as the live freshness view let managers act on aging stock before it expired rather than discovering the loss after the fact. Fill rate reached 96%, up from below-target levels, as automatic replenishment closed the gap between a stock signal and an order being placed.
The combined effect unlocked an estimated Rs. 1.2 crore in margin, split between reduced shrink write-offs and fewer emergency buys made at unfavourable prices under time pressure.
KEEP READING
Explore more studies on AI, analytics, and enterprise intelligence.
