Ai_DL1

A whole team of data scientists, in seconds

AI and machine learning powers hyper-personalization where targeting is done at a true segment-of-one level — the individual customer — leveraging massive amounts of purchase data, calculated brand loyalty, discount propensity, purchase frequency and other categorical trends to suggest the right promotion to the right customer.

AI and Cloud Analytics

More than insights, SMI combines live shopper journey and product data with POS, loyalty, social media and other data sets, to deliver solutions with high impact value. SMI monetizes store data with solutions that enable customer centric merchandising, streamlined operations and personalized customer engagements.

AI and Deep Learning technologies

AI handles fast moving changes in consumer behavior

Retail sensor devices provide live data to the SMI cloud where powerful GPU accelerated computing power enables Deep Learning algorithms to continuously look for patterns to events that occur in your retail spaces.

Instead of waiting for traditional statistical model to be manually updated and maintained, SMI’s AI data analysis is self-learning and is made more accurate as data is collected from a multitude of sources.

Push-intelligence that tells you what you need to know, right now

SMI’s sensor fusion and AI provides a level of diagnostics and insights that uncover what factors are impacting your business, from supply chain to operations, merchandising and sales.

With all your data streams combined, mathematical and forecasting models are able to detect important new conditions that you need to know and sends out notifications to relevant personnel, as soon as event occur.

Monetizing your data with improved operations

A connected retailer realizes productivity gains, from supply chain to merchandising to marketing activities, for example:

Customer Centric Merchandising
Analytics can pinpoint areas of assortment optimization, range localization and better product visibility, resulting in a shopper journey based store layout with improved shopping metrics and return on every square meter of shopping area.

Hyper-Relevant Engagements, by Design
With an ability to know what customers are looking for and need to know at a given moment during the shopping journey, marketers can achieve an unparalleled level of ‘response to conversion’ metrics:

Inventory Efficiencies
IoT powered stores with efficient, battery powered cameras that sends product images to the cloud for product recognition, can provide ongoing information on conditions and predict when shelves need replenishment.

deep_learning
Combine all your data streams together:  live shopper journey and product data with POS, loyalty, social media and other data sets.
CHICAGO - AUGUST 17, 2012:  Record high, drought-driven  corn prices  threaten increases in cereal prices.
A new evolution in CRM manages hyper-relevant and contextual customer interactions, delivers more efficient engagements and offers immediate customer savings.

Cloud Analytics and Solutions

High performance analytical platform

The SMI BI platform components include:

collaborative workspaces with analytical templates, data sources and dashboards that can be packaged as a BI app

interactive visual controls for fast exploration and discovery of data patterns

data management which enables a Data Repository of cleansed and normalized data sources

an analytical processing engine for big data modeling and calculations

Taken together, they constitute the essential requirements for a cloud based BI platform that can deliver solutions to business.

Ready to use analytical models

SMI BI application is presenting ordinary users with a library of re-usable analytical models (templates) that can be applied to the user’s data, thereby eliminating the need for complex and sophisticated programming that can introduce instability or poor performance in case the programmer doesn’t write his code properly.

The innovation is centered on prepackaged, ready to use statistical models that can be applied to any data set:

mathematical models packaged with specially designed algorithms that cluster data into patterns (for revealing undetected dependencies) and predict trends via a variety of regression and time series analysis techniques (e.g. logistic regression, multinomial logit and probit models)

forecasting models that estimate future demands as a function of past historical data, such as last period demand, simple and weighted moving averages (N-Period), simple exponential

Interactive visualisation charts and controls

Together with highly interactive capabilities, SMI users can quickly and easily see patterns, trends, and unforeseen relationships and dependencies in their data – and as a result, users are able to draw insight, inferences, and conclusions that improve performance and provide a competitive advantage.

Dashboards with customizable layout and design features for image background, button styles, legends, chart titles and more

Fast interactive chart controls with intuitive selection of chart areas to select and explore data

Chart and table mash ups in any combination on a dashboard

Changeable measures and drill downs to other dimensions

Matrix chart control for revealing complex data patterns by simply dragging columns and rows to re-order the data according to easy to interpret visual markers

Cloud business applications

SMI offers a range of ready to use business applications that connect all your data sources that stream data into 1 unified repository for access by an application layer with built in workflow that provides solutions that enables customer centric merchandising, streamlines operations and delivers personalized customer engagements.

SMI business applications are fully scaleable and offered on a subscription basis based on the retail environment size and number of stores that are to be covered by the sensor network.

SMI-architecture
matrix
SMI’s patented controls makes visualisation of multi-dimensional data easy to find data patterns
heat_map
Merchandising solutions utilize specialized visualisations such as heat maps for providing a quick read out on conditions

Like to know more?