Let AI summarize this article for you
Umai Retail, the largest FMCG operator in Kyrgyzstan, is introducing new tools in Retano SCM: the Leonardo AI assistant, a control panel for demand forecasts, event and promotion planning, and local assortment management for stores.
About Umai Retail
Umai Retail is part of the Umai Group holding. It operates more than 170 stores under the Globus, Narodny, Dostor and SPAR banners and serves more than 280,000 customers a day.
The company’s supply chain serves very different formats: Dostor hard discounters with around 1,200 items, Narodny neighbourhood stores with more than 7,000, Globus supermarkets and hypermarkets, and the Globus online store. The total range covers around 20,000 items, including fresh products and goods from the company’s own production: a central kitchen, a bakery and in-store production units. Goods reach the stores from nine distribution centres with a total area of more than 23,500 square metres, along more than 40 delivery routes.
Umai Retail has used Retano SCM for demand forecasting and automatic replenishment since 2023. The system forecasts demand from sales history, retrains on new data and generates replenishment orders. As the chain has grown, so has the number of items, promotions and exceptions the supply chain team reviews every day. Checking forecasts across 20,000 items takes more time, demand around local events is not always visible in sales history, and stores need to adjust their range faster than central processes allow.
Leonardo: an AI assistant for supply chain decisions
Leonardo is the AI assistant built into Retano SCM. It explains the system’s recommendations, so a planner can see why a forecast or an order was proposed before accepting it. On request it prepares a daily forecast summary: corrections made, critical tasks pending, anomalies detected, active and upcoming promotions, and forecast accuracy for the previous week. The summary ends with a recommendation on which tasks to handle first and a link to them.
Forecast control panel
In Retano SCM planners work on the forecast itself, not only on the orders it produces. The control panel in the forecast editor shows critical tasks, anomalies detected in the last 24 hours, forecast accuracy, the forecast trend and forecast quality by category, measured by WMAPE and bias. A separate indicator lists products due for delisting and the value at risk.
Events and promotions in the forecast
Some demand peaks have a known cause that sales history treats as noise: stores near a stadium sell more water, beer and snacks on match days. A buyer can load an event calendar, such as a match schedule, for the stores concerned and apply an uplift coefficient to the relevant categories on those dates. All events and promotions are kept in one list with their status, active or upcoming, and can be viewed as cards or as a table. Each entry shows its period, the uplift applied and the number of stores and items it covers.
Local assortment management for stores
Head office sets the assortment matrix for each category, and stores decide what to order within it. Retano SCM gives store managers a tool to manage the local range inside the automatic replenishment system. Category Health, a category-by-store heatmap and gap analysis show categories under pressure, out-of-stock risks, slow-moving items and products that sell well in some stores of a cluster but are missing in others.
The system proposes products to add and to remove, estimates the expected effect on sales and margin, and flags possible cannibalisation. The decision stays with the store manager. Changes go through a controlled workflow with full traceability and can be rolled back. After a change, the system compares expected and actual sales, profit and out-of-stock levels at 30 and 90 days. The tool works alongside central category management and does not replace it.
About Retano SCM
Retano SCM is demand forecasting and automatic replenishment software for retail chains. It forecasts demand from sales history with machine learning models, generates replenishment orders and lets planners and buyers adjust forecasts for events and promotions. The built-in AI assistant Leonardo explains the system’s recommendations.