Analytics Grid
"Retail focused, smart algorithms, high performance."
Array of demand forecasting models
- Multivariate time series models
- Dynamic regression models
- Bayesian models
- Neural Networks
Efficiently manage retail's large scale problems and dimensions
- Very large number of price and non-price variables
- Multilevel distribution channels, store and product hierarchies
- Large set of interdependent and potentially conflicting price positioning rules and performance targets
Comprehensive range of retail specific models
- Demand Forecasting: Daily, weekly, monthly, ...
- Optimising: Prices, promotions, assortments, markdowns, replenishement schedules, safety stocks
- Association Rules learning: Market basket analysis, customers mission profiling
- Clustering & Segmentation: Stores, customers, products
Horizontal and vertical scaling
- Algorithms designed to run in parallel
- Easily implemented on cost-effective public or private cloud based computer services
Data cleaning and analysis
- Completeness, accuracy, relevance
- Distribution analysis, noise evaluation
- Outliers identification and correction