Intelligent Forecasting of Sales and Number of Customers

Challenge:
• • Lack of accurate sales forecasts in the retail and gastronomy industries.
• • Outdated methods based on manual plan creation (e.g. in Excel), not utilizing external factors (e.g. weather, events).
• • Need for data integration and a single, consistent source of information for all departments.

Solution:
• Goal: Forecasting sales, number of customers, and net profit (sales – costs).
• Method: Use of machine learning models (including neural networks and XGBoost) tested on several variations and selecting the most accurate.
• Data: Historical sales results, weather data, marketing activity information.
• End-to-end system:
• Automated data collection and processing in the cloud.
• Interactive Power BI dashboard with clear visualizations.
• Quick deployment for small and medium-sized enterprises.

Benefits:
• • Precise forecasts – accounting for seasonality, promotions, and weather conditions.
• • Time savings – managers no longer spend hours on manual planning.
• • Single source of truth – everyone works with the same, trusted forecasts.
• • Scalability – easy system expansion as data volume grows or the sales network expands.

Results:
• Reduced forecast preparation time from several days to just a few hours.
• Lower risk of overestimating or underestimating demand.
• Better resource planning (e.g. staffing, inventory).

Web Application – Real-Time Operational Dashboard

Challenge:
• Production teams lacked real-time insight into daily sales plans or the number of customers served. They only saw the results at the end of the day, making quick adjustments impossible.

Solution:
• Scope and Data: Sales, number of transactions, average check, service time – all metrics displayed relative to the plan/forecast.
• “Competition” feature: Shows the top three teams, motivating others to improve their performance.
• Real-time: Data is displayed almost instantly, allowing immediate reactions.
• Technology: A web application accessible on TVs in sales units.

Integration and Data Flow:
• Data is sourced directly from sales systems.
• It is processed and aggregated before being displayed.
• The result is presented on a clear dashboard, shown on screens in production areas.

Benefits:
• Immediate action: Teams can quickly address any dip in performance relative to the plan.
• Better coordination: Managers have constant visibility, reducing reaction times and facilitating staff motivation.
• Impact on sales: Dynamically adjusting strategy and resources increases the likelihood of meeting or exceeding goals.

Automatic Employee Scheduling Based on Demand Forecasts

Challenge:
• • Excessive time spent by managers creating schedules manually.
• • Frequent overstaffing, hiring too many employees during low-traffic periods, leading to unnecessary costs.

Solution:
• Utilizing the forecast: The system uses sales and customer forecasts to determine staffing levels needed for each shift in advance.
• Integration with existing tools: Through an API, a grid of recommended positions and hours is sent; managers then assign actual employees.
• Automation: Thanks to intelligent demand planning by hour and day, the system generates schedules up to a month ahead.

Benefits:
• Time savings: Managers no longer spend hours on tedious scheduling tasks.
• Lower costs: Fewer employees are scheduled beyond actual needs.
• High accuracy: Achieves around 90% effectiveness in forecasting demand.
• Clarity and organization: Employees know their schedules in advance, and businesses can better manage resources.

Intelligent Ordering of Semi-Finished Products in Restaurants

Challenge:
• • Inaccurate ordering of semi-finished products (often too many or too few).
• • Wasted unused ingredients and shortages at critical moments.
• • Reliance on staff intuition instead of solid data.

Solution:
• Demand forecasting: Analyzing historical sales, seasonality, and menu changes.
• Recipes and usage: Taking into account precise ingredient usage per recipe (+ a buffer for potential waste).
• Automatic orders: The system generates a list of semi-finished products required to fulfill forecasted demand. The user (e.g. restaurant manager) can review and adjust the proposed order if needed.
• Integration: Linked with weekly restaurant inventories to automatically update stock levels.

Benefits:
• Minimizing waste: Less discarded or expired product.
• Greater availability: Reduced risk of running out of key ingredients, improving service quality and customer satisfaction.
• Time savings: Managers no longer need to guess how much product is required; the system suggests the optimal order list.
• Better organization: Simplifies and standardizes stock planning, leading to smoother operations and lower costs.