
Tills close
Every bill from every till lands within minutes, by store and by SKU. Returns, voids and staff purchases are cleaned out before anything is learned from them.
Demand forecasting + store operations for GCC retail
Tattva forecasts demand for every SKU in every store, every night. Then it turns the numbers into orders, allocations and a morning task list your store teams actually open.
Running every night for 14 retail chains, from 12 stores to 640
01Why stores miss
Gulf demand is local to the street. Ramadan moves eleven days every year. A football final empties one aisle and fills another. A heatwave doubles water sales in one city and halves footfall in the next. Salary week, the summer exodus, school reopening: each store lives on its own clock.
Most chains still plan from a head office average and a spreadsheet. Tattva learns the clock of every single store, then plans the night around it.


02The platform
Start with the forecast, add the modules your teams need. Every element reads the same store by SKU by day plan, so buying, the DC and the shop floor finally work from one number.

01 / Fc
Signals every forecast reads
03Forecast lab
A simplified view of what a planner sees each morning. Pick a store and a category, switch on the events coming up, and compare Tattva with the four week average most chains still order from.
| SKU | On hand | 4 wk avg | Tattva |
|---|
Generated 01:42 GST. Released to DC at 04:10.
04Overnight

Every bill from every till lands within minutes, by store and by SKU. Returns, voids and staff purchases are cleaned out before anything is learned from them.

42 million store by SKU by day forecasts, each with a range, not just a number. Slow movers, new launches and loose staples each get the model that suits them.

Buyers get a kilo level list for the morning market run, split by DC and by grade, with the shelf life of every lot already counted.

Store orders go to the DC rounded to case packs, capped by shelf space and truck capacity. Your ERP receives them as if a planner had typed them.

Routes follow the plan, not habit. Stores that will sell out by noon get the first drop of the day.

The store manager opens six tasks ranked by dirhams at risk. Fill the laban chiller first. Move the Eid packs to the gondola end. Done before the shutter goes up.

The right sizes on the floor, the right quantities in the chiller, and nobody had to guess. Tonight, it starts again with today's bills.
05For the shop floor
Forecasts only matter if someone acts on them at 7 am. The Tattva store app turns the plan into a short, ranked list for each store manager, in Arabic, English, Tagalog and five more languages, on the phone they already carry.
Try it: tick off a task on the phone.
Tuesday huddle
Good morning, Mariam
All done. Shutter up at 08:00.
06Results
Every pilot runs with a control group of stores that keep ordering the old way. These are the numbers after rollout, signed off by the client's finance team.

Wadi Fresh · 212 supermarkets · UAE + Oman

Sahra Modest Wear · 86 stores · Saudi Arabia + UAE

Shifa Plus · 640 pharmacies · Across the GCC
We used to plan Ramadan on a spreadsheet in January and hope. This year Tattva moved stock from Jebel Ali to Al Ain two days before any of us saw the trend.
My store managers stopped calling the DC at seven in the morning. The app tells them what is arriving and what to fix first.
The real test was finance. They now use the Tattva number for the monthly cash plan, and they do not trust anything easily.
07Rollout
We connect to your POS, ERP and WMS through files, APIs or SFTP, and show you exactly where the data is thin.
Tattva forecasts your last two years day by day, blind. You compare it with what you actually ordered and sold.
20 to 40 stores order from Tattva while a matched control group carries on. Results are reviewed weekly with your planners.
Region by region, with your planners in the loop. They approve exceptions; the routine runs itself.
Hosted in UAE data centres. Your data trains your models only.
Single sign on, role based access and a full audit trail of every order change.
Planners can override any number. Tattva learns from the override, and tells you when it was right.
08Company
Tattva is 48 people in Dubai Internet City: former category managers, DC heads and store managers working next to data scientists and engineers. Every engineer spends their first two weeks on a store floor and a night shift at a DC.

Raul MendozaCo-founder. Twelve years running merchandise planning for a Gulf hypermarket chain.raul@tattvaaisystems.store
Amanda PriceCo-founder, Retail Operations. Ran 300 stores and three DCs before Tattva.amanda@tattvaaisystems.store
Steven ClarkeChief Scientist. Builds the forecasting stack and still reviews every new model.steven@tattvaaisystems.store
Tyler HudsonHead of Deployments. Has opened Tattva in more stockrooms than anyone.tyler@tattvaaisystems.store09Questions
No. Most chains have gaps: missing stock counts, merged SKUs, a year of odd promotions. The data audit in week one shows what we can use as it is and what we fix on our side.
Neither. Tattva sits beside your ERP and writes orders into it. Your planners stop typing routine orders and spend their time on exceptions, new launches and suppliers.
New stores borrow the demand shape of similar stores nearby, then switch to their own history as it builds. New products borrow from look alike products by price, pack and category.
Supermarkets and hypermarkets, pharmacy, fashion and modest wear, home and electronics, and quick service food chains. Chains from 12 stores to more than 600.
A fixed fee for the pilot, then a yearly fee per store. Most chains recover the full year in working capital within the first quarter.
10Book a pilot
Tell us about your chain. A retail planner from our team, not a sales script, will write back within one working day and set up a backtest on your own data.