Akashic for retail and consumer goods · Chains, brands & marketplaces
Know what sells. Before it ships.
Demand shows up in your POS in real time and in your planning sheet three weeks later. Akashic connects sales, stock, and supply into one live picture, so the forecast runs on today’s signal, not last month’s export.
Sell-through signal
SKU · store · till time
Stock graph
Shelf · backroom · DC · in transit
Enterprise record
Margin · promo · lead time
Move 240 units · DC-2 → Store 114
Ahead of the weekend peak · trend attached
Sell-through
1,840/hr
Cover
2 days
Stockout
Fri
Systems joined
The shelves were empty. The warehouse was full.
On the shelf
Sold out.
Promotion running · southern region
Shelves empty in one state while the tills keep ringing. The replenishment order hasn’t even been raised.
In the warehouse
Surplus
DC-2 Bhiwandi · 340 units of the same item
POS says one thing, ERP another, the planning sheet a third, and all of them are days old. The replenishment lands after the promotion ends. The markdown follows a month later.
The gap between the shelf and the warehouse is where every markdown is born.
Stock
Shelf
Stock
Backroom
Stock
DC-2
What the live picture is already saying
RESOLVE
One SKU, every location
Shelf, backroom, in-transit, and DC stock resolve into one live position per product.
PROJECT
Stockout seen on Monday
The promotion spike projects a Friday gap while the surplus can still be moved.
ACT
Replenishment routed
Surplus at the northern DC is routed to the stores that need it, before the shelf goes bare.
1
Demand signals arrive late
POS data reaches planners through overnight exports and weekly aggregation. By the time the trend is visible, the moment to act on it has passed.
2
Stock is invisible between systems
The DC, the store backroom, and in-transit inventory live in different systems that reconcile weekly. Nobody holds one live picture of where product actually is.
3
The forecast lives in a spreadsheet
Planning runs in files detached from live sales, weather, and promotions. Every forecast is stale on arrival and adjusted by gut feel.
None of these are people problems. All of them are architecture problems.
Every empty shelf is a promise you didn't get to keep.
Scroll through your own chain: the players you already have, and what changes when their data lands on one governed record.
- 01
The players
Every link keeps its own count.
Stores, e-commerce, the DC, logistics, HQ planning: each holds its own number for the same product.
- 02
Tills stream
Every sale is a live signal.
POS and e-commerce sales stream in as tills close, not through Monday's export.
- 03
Stock resolves
One picture, DC to shelf.
Warehouse, in-transit and backroom stock resolve into one live position per SKU, per store.
- 04
The gap closes
Stock moves before the shelf empties.
The Friday stockout is projected on Monday morning, and 240 units are already on the road.
One live picture, from shelf to supplier.
Akashic unifies POS, ERP, warehouse, and supplier feeds as they change, resolves every product and location into one record, and puts the forecast on live signal instead of last month's export.
01 / Unify
Every till, one stream
Data Pipelines connect POS, e-commerce, ERP, WMS, and supplier EDI in real time. Master Data resolves every SKU, store, and supplier into one governed record.
02 / Govern
One version of stock
The warehouse holds sales, inventory, and movement history in governed layers. When merchandising and finance pull a number, it's the same number.
03 / Predict
Forecast per store, not per region
Models trained on your own sales, promotions, and seasonality forecast demand at store-SKU level, and flag supply anomalies before they become empty shelves.
04 / Act
Replenishment answers, not reports
Planners ask which stores run out this week and get an answer with the stock, the trend, and the source attached. Merchandising watches live sell-through, not the Monday export.
Every module runs on the same governed foundation · every answer arrives with its lineage
What changes in the chain.
Same stores. Same suppliers. Decisions made on today's demand instead of last month's.
BeforeDemand visible in weekly aggregated exports
AfterLive sell-through, per store, per SKU
BeforeStock reconciled between systems weekly
AfterOne live picture from DC to shelf
BeforeForecast adjusted by gut feel in a spreadsheet
AfterStore-level forecast on live signals
BeforeMarkdowns absorbing every planning miss
AfterSurplus flagged while it can still move
Let's talk about your demand signal.
Whether it's stockouts your exports catch too late or a planning cycle that runs on stale spreadsheets, we'd like to show you what forecasting on live, governed data looks like.