Turn Data Chaos into Data Value

At a Glance

360-Degree
Crime Analysis

Reduce crime using intelligent data

AT A GLANCE

A North American city wanted to lower crime rates and create a safer environment for citizens. Law enforcement and social services organizations had collected huge volumes of data from a variety of sources. The city hoped to use this data to better understand the causes of crime, gain insight from relevant correlations, and predict the impact of crime on citizens.

Challenges

Challenges

Multiple municipal systems for crime reports, including 911 calls, police reports and actions, and social network feeds
Lack of understanding about how buildings and infrastructure affect crime
Inability to gain insight from additional data points – such as socioeconomic data, weather data, and local and global events

Solution

Solution

Machine-learning clustering algorithms that increase understanding of past crimes and help predict future crimes
Multidimensional analysis, which helps analysts discover meaningful correlations
Predictive models that help analysts analyze and predict crime

Benefits

Benefits

Increased insight into historical crime
Enhanced ability to prevent crime based on predictive capabilities
Greater understanding of how crime affects citizens and communities

Dig Deeper

Dig Deeper

Read this blog to learn about SAP Data Intelligence.

At a Glance

360-Degree
Customer View

Understand and anticipate consumer requirements

AT A GLANCE

A global footwear and sports equipment retailer wanted to become a consumer-centric business to support its long-term growth. To better understand customer needs and predict demand, the company needed a better way to gain insight from its sales and tracking data (see Figure).

Challenges

Challenges

Lack of a single consolidated view of the consumer
No instant access to customer-centric decision-support data
Siloed data, with consumer transaction history in on-premise enterprise systems and real-time data in the cloud

Solution

Solution

Integration of and access to on-premise and cloud landscapes without data duplication (see Figure)
Orchestration and push-down features for combining individual data sets

A working environment designed for data scientists, supporting Python and R integration, for automated model creation and access
End-to-end view of data that captures everything companies need to know about consumers

Benefits

Benefits

Increased business value of data, with better insights gained from running analytics on complete, trusted data from across the enterprise
Enhanced ability to create predictive insights about consumers, with simplified tasks for data scientists
Reduced time to value for data science projects
Improved data consistency, governance, and transparency

Dig Deeper

Dig Deeper

Learn more about data orchestration in this O’Reilly report.

At a Glance

Cognitive Customer Service

AT A GLANCE

In the utilities industry, the volume of calls received by call centers is extremely high and leads to delays in resolution, affecting customer satisfaction. Cognitive customer service applies machine learning to predict the reason for a customer call based on the pattern of past communication. It’s able to propose the top three “next best actions” that the call center executive can choose to recommend to the customer, which can bring down the overall time spent on call resolution and improves customer satisfaction.

Challenges

Challenges

Customer call center representatives are challenged by the inability to:

Quickly understand why a customer called
Provide personalized and proactive service while reducing incorrect and inconsistent service
Predict call volumes
Reduce manual and mundane activities during a customer interaction

Solution

Solution

Using machine learning, call centers are able to:

Leverage customer information and interaction history to suggest possible context for a customer call
Use SAP Data Intelligence to extract relevant demographic and interaction data from SAP S/4HANA and the model trained and deployed to enable inference
Use predetermined rules and operating procedures, based on the reason for a call, to suggest the next best actions
Benefit from the solution, which is integrated with the customer interaction center in the CRM system

Benefits

Benefits

Machine learning model to predict the reason for a customer call and propose personalized next best actions based on various influencing parameters and interaction history
Collection of REST APIs, developed using SAP Data Services software, and open APIs to gather insight from the customer from SAP master data, transactional data, and the Internet
Collection of attended and unattended intelligent robotic process automation applications from SAP to automate a set of actions
Intuitive SAP Fiori app that can be embedded in any of the CRM applications

*SAP HANA database modeling with more than 25 configuration tables and more than 10 master and transactional tables

Dig Deeper

Dig Deeper

Read the blog or watch the video to learn more.

Scenario content provided by Wipro

At a Glance

Commodity Code Prediction

Identify and assign the proper export values

AT A GLANCE

A global manufacturer produces goods in one country and sells them around the world. By law, the manufacturer must assign export codes to these goods – a time-consuming, manual effort that often requires expert intervention (see Figure).

Challenges

Challenges

Mandatory import and export rules requiring assignment of commodity codes
Product delays at the border when codes are missing or incorrect
Inconsistent master data that increases the need for manual code processing
Elevated organizational costs for export activities

Solution

Solution

Machine learning models that can predict the correct value for each field (see Figure)
Automated learning of logic behind complex fields (such as the commodity code)
Understanding of dependencies for various data attributes

Benefits

Benefits

Reduced manual effort for commodity code assignment
Increased validity of assigned commodity codes
Improved customer satisfaction due to export compliance
Reduced fees and customs-related noncompliance issues

Dig Deeper

Dig Deeper

Learn more about machine learning with SAP Data Intelligence.

At a Glance

Asset Data Intelligence

Increase safety and reduce regulatory risk by improving core asset master data

AT A GLANCE

A utilities company wanted to accelerate efforts to improve the quality of their core asset master data to enable data-driven asset management, increase safety, and reduce regulatory risk. With asset data intelligence, it has the AI technology to autonomously find patterns and outliers in asset data – reducing the time and investment required to identify quality issues. Supervised data entry exposes issues to users in real time and prevents data issues directly in the ERP system. Analytical dashboards identify overarching data issues, allowing effective and efficient remediation of problems at scale.

Challenges

Challenges

Poor quality asset data that impacts safety and reliability and increases regulatory risk
Unsupervised manual data entry that causes systematic data quality issues

Traditional methods for finding data issues, resulting in:
- High effort to manually define complex relationships to support rule-based validation
- Reliance on users to identify and raise data concerns
- Unmitigated business risk around unidentified issues

Solution

Solution

This end-to-end, automated solution supports proactive and reactive asset data governance.

Explainable AI that presents results to users with a narrative explanation of the context of the outliers found and their driving characteristics

Analysis and improvement of overall data quality, with results that are published and visualized in SAP Analytics Cloud solution dashboards (reactive data governance)
Root-cause analysis of poor data quality through the creation of OData services on SAP Business Technology Platform that offer real-time identification, prevention, and correction of outliers in source applications (proactive data governance)

Benefits

Benefits

Improve the quality of core-asset master data to increase accuracy of analysis and operational decisions
Prevent quality issues at the source to reduce the impact of bad data permeating throughout the organization
Increase accuracy for asset health monitoring and reduce operational and regulatory risk
Accelerate time-to-value with an automated implementation (as compared to traditional rule-based and manual approaches)

Dig Deeper

Dig Deeper

Read the blog to learn more.

Scenario content provided by EY AgilityWorks

At a Glance

Customer Churn Prediction

Increase customer retention

AT A GLANCE

A major chemical company wanted to identify the reasons for high customer churn rates and determine the actions it should take to retain customers. With sales orders and customer purchasing history in separate IT systems, the company struggled to understand customer motivations or predict defections (see Figure).

Challenges

Challenges

Siloed data and nonintegrated business systems
Complex lifecycle management of current solutions
Limited reusability and modularity of existing machine learning processes

Solution

Solution

Centralized data management environment, supporting management and orchestration of data sources and execution of machine learning algorithms (see Figure 10)
Flexible integration with machine learning libraries, allowing the company to use a variety of algorithms
Reusable data flows that can be applied to different scenarios

Benefits

Benefits

Improved understanding of customer behavior for both churn and repeat-buy scenarios
Enhanced ability to quickly predict and adjust to customer actions
Increased customer retention by proactively reaching out to customers and offering promotions to avoid churn

Dig Deeper

Dig Deeper

Learn more about data orchestration in this O’Reilly report.

At a Glance

Customer Risk

Protect enterprise sales with insight into partner financials

AT A GLANCE

To better understand the likelihood of being paid by its supply chain partners, a manufacturer wanted to assess business, financial, and sentiment data to gain real-time insights. By more effectively scoring the risk of doing business with each partner, the company could make better credit decisions that would improve the bottom line (see Figure).

Challenges

Challenges

Difficulty collecting and analyzing huge volumes of Big Data
Lack of timely insight to inform credit decisions
Elevated risk of delayed payment or default by supply chain partners

Solution

Solution

Applied intelligence that combines credit management data, business partner master data, and sentiment analysis of social media feeds (see Figure)
Risk-scoring model powered by a machine learning–driven algorithm

Analytical dashboard to support intuitive, risk-based decision-making
Addition of a custom check in the sales process that requires additional approvals for sales orders that exceed safety thresholds

Benefits

Benefits

Increased understanding of customer financial conditions and pressures
Faster, more complete insight into customer payment risk
Reduced number of late payments, improving financial outcomes

Dig Deeper

Dig Deeper

Take a guided demo of SAP Data Intelligence.

At a Glance

Digital
Enterprise Asset Management

Extend the life of complex assets

AT A GLANCE

A leading utility wanted to better protect the efficiency, reliability, and health of costly assets such as wind turbines and power lines. The company collected extensive maintenance-record data and maintained-asset master data in its business systems. It also streamed sensor data from each asset. Yet pulling this information together to gain real insight seemed impossible (see Figure).

Challenges

Challenges

High cost of asset maintenance and repair
Multiple discrete sources of data for each asset
Lack of unified insight into asset health and management requirements
Need for predictive information to reduce asset downtime

Solution

Solution

SAP Data Intelligence as the foundation for integrating, orchestrating, and processing data (see Figure)
Advanced predictive maintenance models that use machine learning techniques

Simulation of physical characteristics of each asset, creating a digital twin
Smart applications such as predictive engineering insights, failure prediction, and anomaly detection

Benefits

Benefits

Simplified, streamlined data management for all data related to the asset’s maintenance and health
Extended efficiency, reliability, and health of complex assets
Increased understanding of assets through predictive maintenance and engineering insights

Dig Deeper

Dig Deeper

Listen to Episode 21 of the Let’s Talk Data podcast: Modernizing Analytics with Machine Learning.

At a Glance

E-Commerce Product Placement Optimization

Deliver consistent product information across online platforms

AT A GLANCE

The account management team at a beverage company needed a better way to help ensure consistent product pricing, placement, and messaging on more than 100,000 third-party e-commerce platforms. Instead of manually scouring Web sites and checking products one by one, the team wanted to automate the process so it could rapidly detect errors and suggest corrections – at scale (see Figure ).

Challenges

Challenges

Partnership with a large number of third-party e-commerce sites, each with vastly different formats and layouts
Time-consuming manual effort required to check e-commerce detail to ensure compliance with product guidelines
Difficulty scaling manual processes to run accurately and efficiently

Solution

Solution

Intelligent application that reuses the company’s Web-crawling tools to extract product images and text from e-commerce platforms and applies computer vision to identify and classify product images (see Figure 16)
Automated data cleansing and preparation of unstructured data

Streamlined comparison of images to product catalog photos and comparison of e-commerce data with master data
Generation of recommended corrective actions
Model lifecycle management to keep current with the changing product catalog

Benefits

Benefits

Reduced time to detect and correct outdated or false information on e-commerce platforms
Improved scalability of the optimization process across hundreds of thousands of e-commerce platforms
Superior brand and product management through automation

Dig Deeper

Dig Deeper

Listen to Episode 12 of the Let’s Talk Data podcast: Challenges of Deploying Machine Learning.

At a Glance

HR Skills
Matching for
Open Positions

Quickly find the most qualified workers for internal job postings

AT A GLANCE

A large bank wanted to publicize job openings to employees first. By giving internal workers this head start, the bank hoped to fast-track employees for more advanced roles and promotions. Yet the bank’s HR team struggled to find, sort, and analyze employee skills information – making it difficult to identify and promote the best candidates.

Challenges

Challenges

Widespread, nonintegrated storage of employee information – including skills, locations, and scoring data – across multiple applications and spreadsheets (see Figure)
Manual effort required to collect and assemble data from multiple sources

Inability to proactively notify employees of available positions that match their skills and support career growth
Ineffective talent- and employee-retention efforts

Solution

Solution

Integrated employee information from multiple sources (see Figure)
Machine learning capabilities in the R programming language that find and score candidates for job fitness
Real-time matching of employees to fill internal positions

Benefits

Benefits

Reduced need to hire and train external candidates
Improved employee retention and employee engagement
Increased cost savings for external recruitment

Dig Deeper

Dig Deeper

Read the Forrester report on the total economic impact offered by machine learning features in SAP solutions.

At a Glance

Intelligent Rule Mining

Analyze data dependencies and improve overall data quality in your enterprise

AT A GLANCE

A global company wanted to become more data driven by identifying data dependencies across enterprise applications. Its goal was to identify data inconsistencies and improve overall data quality.

The company used data rules, which describe business operations, definitions, and constraints that apply to an organization’s data. Data rules reflect the business structure, serve to guide or control organizational behavior, and are put in place to help an organization achieve its goals. The company’s goal was to mine the data rules by automating the extraction of complex data dependencies, check their applicability across several systems, and drive data quality and data-driven business decisions.

Challenges

Challenges

Identification of incorrect data entries as a complex and time-consuming task performed manually by data stewards or data owners in enterprises
Difficulty checking data rule applicability across several systems

Solution

Solution

Rule-mining operator who takes the data as an input and outputs the mined results to be stored in the knowledge base
Operator leveraging a local Apache Spark cluster to enable distributed computing capabilities for mining rules, allowing faster processing times

Data provisioned from different clouds or on-premise databases (MSSQL, MySQL, and so on)
Data dependencies identified and improvement in the overall data quality of your enterprise

Benefits

Benefits

Improved data quality and efficiency and stronger decision-making in your enterprise
Ability to identify and compare complex enterprise rule sets against additional data sources within your enterprise
Detection of data inconsistencies through outlier detection, confidence, and frequency distributions with input suggestions and automated data population options

Dig Deeper

Dig Deeper

Read to blog or watch the video to learn more.

Scenario content provided by Camelot ITLab

At a Glance

Internal
Auditing

Revolutionize auditing analysis

AT A GLANCE

A large state government wanted to modernize its internal auditing processes for financial transactions. With a budget of $85+ billion, the agency hoped to improve its analysis and reduce risk by increasing insight into spending.

Challenges

Challenges

Multiple data silos for financial transactions, including data about procurement cards, grants, and contracts
Lack of a single view into total spend across multiple data sources
Significant manual effort required to enable internal auditing
Inability to conduct proactive audits based on intelligent insights

Solution

Solution

Machine learning features that create a single view of global spend
Analytical risk-based models for audit process that offer insights into global spend and risk areas
AI-enabled solution that supports a review of all processes (not just sample data), resulting in the identification of exceptions and critical processes requiring team review

Benefits

Benefits

Single view of global spend across the enterprise
Modernized audit process, which proactively identifies audit candidates with qualitative risk scores
Greater insight into potentially risky spend across millions of transactions and years of data
Simplified implementation of action plan to reduce compliance risk

Dig Deeper

Dig Deeper

See how an Innovation Award winner revolutionized internal auditing.

At a Glance

IoT Data Analysis and Management

Boost performance with real-time, sensor-data analysis

AT A GLANCE

To ensure its wind turbines operate at maximum efficiency, a utility wanted to combine IoT sensor data – such as wind speed, rotor speed, gear-oil temperature, and power average – with historical data. The company hoped to use this information to predict turbine conditions and make performance-enhancing adjustments (see Figure).

Challenges

Challenges

Need for constant monitoring of IoT data
Lack of integration between sensor data and enterprise data
Inability to use data-driven insights to predict future performance or optimize turbine efficiency

Solution

Solution

Integration between turbine sensor data and third-party data sources, such as weather, market, and vendor information (see Figure)
Real-time reporting based on sensor data
Connectivity across the infrastructure, including cloud, on-premise, and edge solutions

Benefits

Benefits

Reduced costs due to higher turbine uptimes and less-frequent need for maintenance
Improved safety of turbine operations
Optimized resources through more accurate performance predictions
Foundation for future, unsupervised machine learning by turbines

Dig Deeper

Dig Deeper

Take a guided demo of SAP Data Intelligence.

At a Glance

Late-Payment Risk Assessment

Get paid faster, with less effort

AT A GLANCE

To improve its financial stability, a consumer goods company wanted to reduce the number of late customer payments. By better understanding its payment risks, the firm hoped to improve cash flow, reduce the time spent on debt collection, and decrease the need for income write-offs (see Figure).

Challenges

Challenges

Compromised cash flow due to late customer payments, despite strong revenues
Inaccurate working capital calculations and cash-flow forecasts
Time-consuming debt-collection activities and bookkeeping tasks to account for write-downs, write-offs, and revenue reversals

Solution

Solution

Ability to merge integrated ERP and financial data with historical payments, collections data, and data from external sources (see Figure)
Modeling capabilities that assess when and if payments are likely to be made, based on historical performance and economic indicators

Scoring technology to rate new accounts receivable, predicting payment dates and the risk of nonpayment

Benefits

Benefits

Increased accuracy of predictions for better projections of working capital and cash flow from sales
Maximized cash collection, reducing days sales outstanding and write-offs for bad debt
Improved relationships by proactively altering credit terms for customers unlikely to meet existing payment terms
Enhanced customer loyalty by rewarding good payment behavior

Dig Deeper

Dig Deeper

Learn more about data intelligence.

At a Glance

Master Data
Creation for
Complex Materials

Automate and accelerate materials creation

AT A GLANCE

A leading provider of specialty chemicals wanted to streamline the management of complex materials data. Creating new specialty chemical products in the manufacturer’s ERP systems relied on complex data structures, multiple types of documents, and error-prone manual data entry. The company wanted to automate the creation and handling of complex materials.

Challenges

Challenges

Complex classification requirements for chemicals, including storage requirements, complex labeling, and safety and environmental classifications (see Figure 27)
Reliance on deep domain expertise
Use of unstructured documents such as images, PDFs, word processing documents, and blueprints

Intensive manual data entry to ensure new materials are created correctly
Incorrect packaging and out-of-stock products due to erroneous material information

Solution

Solution

Machine learning embedded in ERP solution to gather key materials characteristics
Computer vision used to extract key facts from packaging specification documents
Automatic material classification for streamlined management of materials data

Benefits

Benefits

Improved data quality, thanks to less manual data entry
Accelerated processing of complex packaging specification information from vendors
Reduced out-of-stock situations with more accurate orders based on correct packaging specifications
Additional time for staff to focus on high-value activities

Dig Deeper

Dig Deeper

View an example of intelligence automations to streamline the handling of complex materials data. Read this blog to learn more.

Scenario content provided by Camelot ITLab.

At a Glance

Medical Supply
Ordering

Save patient lives with data

AT A GLANCE

Seeking new ways to improve treatments, outcomes, and healthcare experiences, a German distributor of medical products wanted to automate the process of replenishing stock in hospitals and operating rooms. Clinicians spent too much time ordering, tracking, and replenishing medical supplies. These manual processes contributed to poor process transparency and elevated operational costs.

Challenges

Challenges

Excessive manual effort by clinicians to reorder essential medical products (see Figure)
Difficulty analyzing data across multiple internal and external sources
Lack of consistent access by medical personnel to needed supplies
Subpar patient care, due to clinicians spending valuable time checking and ordering supplies

Solution

Solution

One common data model, with integrated data from multiple sources (see Figure)
Enriched data sources, including internal sales and planning information, Google trends, Twitter posts, weather data, and institutional information

Analytics technology to better understand customer challenges
Use of innovative technologies such as IoT, AI, and predictive analytics in sensor boxes, to automatically replenish inventory

Benefits

Benefits

Optimized manufacturing processes and supply chains using intelligent data
Guaranteed availability of supplies, exactly when healthcare providers need them
Reduced manual effort for ordering and tracking supplies
Improved process transparency and reduced operational costs through data orchestration
Increased sales revenue using a new data-driven inventory management solution

Dig Deeper

Dig Deeper

Read more about data management and analytics for healthcare.

At a Glance

Mineral Processing
Plant Productivity

Boost asset utilization for mineral processing with real-time data analysis

AT A GLANCE

To operate its mineral processing plant at maximum efficiency and safety, a mining company combines equipment IoT sensor data with throughput and maintenance data. The company wanted to streamline maintenance shutdown schedules, optimize the remaining useful life of equipment, and reduce the number of emergency shutdowns.

Challenges

Challenges

Improved safety of mining sites, reduced operational costs of asset utilization, decreased number of outages, optimized maintenance, and reduced emergency stoppages due to equipment wear (see Figure)
Need for actionable insights, such as predictive maintenance for high-value mining assets, derived from interpretation of amalgamated data

Operations in extremely distributed environments, including mining sites and oil platforms, where real-time data insights and immediate actions are required
Need to collect data at the edge in real time

Solution

Solution

Asset information management and analytics solution to collect IoT, computer vision, and light detection and ranging (Lidar) data from equipment, as well as asset management data (see Figure)
Real-time insights and accurate predictions
Collection and processing of computer vision and Lidar data at the mineral processing site

Real-time analysis of sensor data, supporting immediate decision-making in the field
Predictive streamlining of equipment maintenance, minimizing shutdowns and optimizing maintenance schedules

Benefits

Benefits

Timely replacement of critical components
Reduced outages at mineral processing plants
12% reduction in maintenance, thanks to a decrease in unscheduled stoppages and safety incidents due to premature equipment wear
10% revenue increase through increased production by monitoring equipment wear and tear and providing maintenance alerts

Dig Deeper

Dig Deeper

Listen to Episode 40 of the Let’s Talk Data podcast podcast and read the blog.

Scenario content brought to you by Intel.

At a Glance

Mobile Asset Productivity and Safety

Boost mining asset safety and efficiency with real-time data analysis

AT A GLANCE

To ensure its mobile assets operate at maximum efficiency, a mining company required greater insights into the operations of its assets. To improve operational safety, reduce asset downtime, and increase productivity, the company wanted to combine computer vision with IoT sensor output, such as anticollision Lidar, GPS, tire pressure, and temperature data.

Challenges

Challenges

Risk of collisions and safety incidents (see Figure 33)
Costly outages due to asset issues, such as tire wear, loss of pressure, and overheating
Operations in extreme environments, such as open-cut or underground mining sites
Need to collect data at the edge in real time to support timely actions

Solution

Solution

Asset information management and analytics solution to collect IoT data, computer vision data from various equipment types, and asset management data for real-time insights and accurate predictions (see Figure)
Integration of streaming data from IoT equipment, computer vision technology, and ERP asset data
Processing of computer vision data in vehicles, at the site

Real-time analysis of sensor data, supporting immediate decision-making in the field
Predictive streamlining of equipment maintenance
Integration of vehicle asset management data and external data sources, such as weather, vendor, and market information

Benefits

Benefits

Up to 20% reduction in tire-related outages, creating multimillion-dollar savings
Up to 30% reduction in collision incidents, improving safety and lowering outage costs
10% decrease in unscheduled stoppages due to premature equipment wear
Optimized resources through more accurate performance predictions

Dig Deeper

Dig Deeper

Listen to Episode 40 of the Let’s Talk Data podcast and read the blog.

Scenario content brought to you by Intel.

At a Glance

Next-Best Decision-Making

Automate and accelerate decision-making

AT A GLANCE

A global insurance company wanted to automate the decision-making process for creating insurance quotes, assessing customer risk according to their individual characteristics. Therefore, the company wanted to facilitate the usage of historical and current data to gain more actionable insights. The company hoped this automation would provide a faster, more efficient, and more precise process for providing individually priced insurance products based on customer input.

Challenges

Challenges

Manual decisions of employees, including the risk for human errors and emotions, mainly driving enterprises
Cost-intensive, time-consuming, and limited scaling of manual decisions due to computational power
Inefficient and ineffective decision-making creating high risk in business processes with a potential negative impact on the corporate bottom line

Solution

Solution

Manual decisions replaced with automation leveraging existing enterprise data and advanced analytics
Use of innovative cloud technology with a flexible and highly scalable architecture
Automated evaluation of customer input enabled using predictive analytics and a proposal of personalized insurance products

Benefits

Benefits

Reduction of cost- and time-intensive manual processes leading to additional availability of human capital
More efficient and accurate provision of insurance quotes by leveraging machine learning
Scalable and flexible architecture by using cloud technology
Elimination of risk from a biased or emotional decision-making process

Dig Deeper

Dig Deeper

Watch the video and read the blog to learn more.

Scenario content provided by Deloitte

At a Glance

Optimizing Asset Effectiveness

Predict asset downtime

AT A GLANCE

An OEM manufacturer has a business objective to increase overall equipment effectiveness (OEE) by reducing unplanned downtime. Unplanned downtime impacts not only the amount and quality of products being produced, but it can also result in significant rework due to lack of effective equipment.

The main impacting factor on OEE is unscheduled equipment downtime for technical reasons or as an operational consequence of another type of downtime. The move to predictive maintenance will allow maintenance engineers to shift their focus on achieving a higher equipment uptime figure, reducing scheduled maintenance to tasks designed to increase uptime based on early warnings for production machines and processes.

Challenges

Challenges

Unscheduled downtime resulting in product deficiencies
Quality loss and speed loss impacting OEE
Scheduled maintenance tasks not linked to actual equipment effectiveness

Solution

Solution

Apply data preparation and machine learning algorithms to:

Understand outliers and anomalies that impact or cause downtime
Identify patterns to call out potential sequencing based on whether a failure will happen
Predict a failure based on the combination of anomalous records and the transaction history

Benefits

Benefits

KPIs such as health scores, top-performing assets, and life expectancy for assets
Reduced operational risk of mission-critical equipment
Increased ROI on assets by predicting failures before they occur
Real-time maintenance, potentially lowering costs
Lowered customer attrition, improved brand image and lost sales
Patterns connected to various maintenance problems
Enablement of customer retention
Just-in-time inventory enabled by estimating order dates for replacement of parts through SAP Ariba solutions

Dig Deeper

Dig Deeper

Read the blog to learn more.

Scenario content provided by IBM

At a Glance

Optimizing Manufacturing Inventory

Balance future demand and safety stock

AT A GLANCE

An IT device manufacturer wanted to ensure inventory levels would enable seamless production without wasting money on excess inventory. Through the creation of a parameterized and automated replenishment policy, the company wanted to balance having enough parts for a seamless production without spending too much money and effort on excess inventory. It also wanted to optimize bills of materials (BOMs) across different product sets to be able to buy parts at larger economies of scale (see Figure).

Challenges

Challenges

Lack of a global view or automated predictions of what inventory levels should be
Demand forecast that creates either too much inventory or inventory shortages
Stock replenishment requiring significant manual effort
Unbalanced inventory levels that drive lower customer service

Solution

Solution

Advanced demand forecast based on machine learning
Advanced predictions providing an ongoing, well-balanced inventory level
Optimized bill of materials across product sets, which increases purchasing power
Automated and customizable replenishment policy
Smart inventory management application that minimizes overall manual effort

Benefits

Benefits

Balanced inventory for production line
Significant cost savings from automated stock replenishment
Increased customer satisfaction
Highly automatized and customizable solution based on a software process with little manual intervention
Substantially higher precision due to markedly minimized calculation intervals
Reduced time to value

Dig Deeper

Dig Deeper

Read the blog or watch the video to learn more.

Scenario content provided by Lenovo

At a Glance

Optimizing Order Fulfillment

Predict order delivery risk

AT A GLANCE

A large wholesale distributor wanted to reduce product delivery delays and supply shortages. It wanted to be able to identify the root causes of delivery delays so it could improve customer loyalty, mitigate the risk of customer attrition due to recurring service-level defaults, and avoid penalties for not meeting service-level agreements while maximizing revenue recognition potential.

Challenges

Challenges

Inability to fill all orders because of insufficient product
Issues predicting staff required to fulfill orders in the distribution center
Difficulty managing risk of delayed shipments and late payments
Need for machine learning to provide 360-degree view of data
Lack of ability to go beyond dashboard to understand why issues exist and take action to mitigate risk

Solution

Solution

360-degree view of data passed through a machine learning algorithm to create order delay likelihood
Understanding of relationships that tell you why an order is late
Prediction of how many days an order will be late
Prioritized orders based on revenue risk and expected delivery delay

Benefits

Benefits

Improved inventory replenishment
Optimized logistics and workforce planning
Mitigated revenue recognition risk
Enhanced customer satisfaction
Mitigated supplier and partner delivery risk

Dig Deeper

Dig Deeper

Read the blog or find the solution on SAP Store to learn more.

Scenario content provided by Inspired Intellect

At a Glance

Personalized
Financial Services
Marketing

Boost sales of financial products and reduce customer churn

AT A GLANCE

A leading financial services firm wanted to better understand the financial position of its customers so it could offer targeted offers. By developing insight into customers’ daily liquidity forecast, the bank expected to better understand client needs and improve the customer experience, increasing satisfaction.

Challenges

Challenges

Inability to gain a clear view of each customer’s liquidity without timely updates from sources such as account transactions and credit card transactions (see Figure)
Difficulty providing on-the-fly offers to match competitor bids, as discovered by customers

History of offers that failed to match customers’ current liquidity, causing clients to distrust bank recommendations
Difficulty operationalizing a machine learning model that accurately predicts the next-best offer for the customer

Solution

Solution

Machine learning models, which can be trained and automatically validate the daily liquidity model (see Figure)
Automated, on-demand retraining of machine learning models
Self-service analytics and responsive dashboards for faster data understanding

Benefits

Benefits

Increased revenue and higher customer satisfaction by offering the right products and services
Closer match between offers and daily customer liquidity, thanks to constantly trained and retrained machine learning models
Self-service analytics for faster data understanding

Dig Deeper

Dig Deeper

Read the analyst reports.

At a Glance

Predictive Quality

Instantly spot and correct manufacturing issues

AT A GLANCE

To improve production quality and lower costs, an automobile company installed heat and pressure sensors, infrared cameras, and temperature gauges on an assembly line for molded metal parts. Despite the huge volumes of data generated by these technologies, the company struggled to detect production problems before full batches of defective parts rolled off the line (see Figure).

Challenges

Challenges

Fragmented data from multiple sources, including enterprise business systems
Lack of timely insight into manufacturing quality issues
Increased material scrap, raising production costs

Solution

Solution

Advanced Internet of Things (IoT) ingestion and orchestration technology that delivers detailed data analysis through a single process (see Figure)
Integration of data from sensors and infrared cameras with logistics data from enterprise systems
Statistical algorithms to calculate quality parameters, describing local and global heat distributions

Machine learning–based information extraction from images
Smart quality management application that uses intelligent processing to deliver instant insight that improves outcomes

Benefits

Benefits

Significant production cost savings
Greater understanding of manufacturing processes and need for adjustments
Improved performance of predictive quality models
Enhanced ability to detect maintenance requirements
Increased agility in complex IT and manufacturing landscapes

Dig Deeper

Dig Deeper

Learn more about data orchestration in this O’Reilly report.

At a Glance

Predictive Simulation

Use data-driven insights to reinvent your business

AT A GLANCE

To satisfy and retain its largest customers, a leading utility wanted to offer value-added services beyond pure grid operations and maintenance. To help municipalities become green cities, the utility intended to turn its data into insights and provide advisory services on the best investment options and infrastructure needed to produce renewable energy (see Figure).

Challenges

Challenges

Multiple diverse, nonintegrated data sources needed to create a complete picture
Data required for analytics and services distributed across multiple systems
Difficulty creating and maintaining a unified view of current data

Solution

Solution

Connectivity among multiple data sources, including energy production, consumption patterns, assets and capacity, grid load, energy pricing, and weather (see Figure)
Orchestration of data pipelines and integration of machine learning capabilities to create a unified view of diverse data with minimum data movement

Monitoring of data flows across a diverse landscape of IT systems
Interactive end-user interfaces for simplified data access

Benefits

Benefits

Increased opportunity to offer new value-added services by developing strategic partnerships with customers
Enhanced ability to benchmark customer data against averages and peers
New simulation functionalities to assess potential investments in renewable energy
Accelerated time to build and access unified data views
Greater agility in applying data science to huge volumes of data
Reduced time to value for data

Dig Deeper

Dig Deeper

Learn more about data orchestration in this O’Reilly report.

At a Glance

Product Quality Improvement

Gain insight from sensor data

AT A GLANCE

A manufacturer of home appliances wanted to gain more insights from its products’ sensor and machine data as well as IoT data about user behavior. Yet fragmented data sources prevented the company from performing product quality analysis, executing maintenance activities, making design changes, or improving customer engagement (see Figure).

Challenges

Challenges

Reliance on extensive programming and scripting efforts to integrate disparate data from various sources
Multiple, individual tools and manual steps required for data integration and coding
Lack of end-to-end monitoring, scheduling, and automation

Solution

Solution

Automated management and orchestration of data processes, from data ingestion to enterprise applications, visualization, and analytics (see Figure 52)
Centralized development tool that offers visual modeling

Integrated machine learning data models
Unified metadata cataloging and governance

Benefits

Benefits

Significant reduction in manual effort, thanks to automation of data flow modeling, execution, monitoring, and management processes
Enhanced ability to thoroughly analyze customer behavior and gain valuable insights into product usage
Faster decision-making with lower costs
Improved performance of predictive models

Dig Deeper

Dig Deeper

Listen to the podcast on The Role of DataOps in Supply Chain.

At a Glance

Sales Activities Value Prediction

Spend time on highest-value activities

AT A GLANCE

A consumer goods company wanted to ensure the sales team was spending time on sales activities with the highest potential for monetary value. While it has activity reporting that enables it to evaluate past activities, the company wanted to predict which future activities could benefit the company the most, to optimize time spent by the sales team.

Challenges

Challenges

Sales teams covering many accounts and having many activities going on simultaneously
Sales teams investing time in activities that do not always have high-yield results
Sales teams wanting to focus their time on activities that will provide the highest monetary value

Solution

Solution

Analysis of all activities in the SAP Sales Cloud solution and SAP CRM
Ability to combine with net sales revenues from SAP S/4HANA
Application of prediction models on past activities and sales to predict which future activities have the highest monetary value
Creation of a dashboard in SAP Analytics Cloud to apply predictions and simulations to future activities

Benefits

Benefits

Understanding of which activities are most likely to yield the highest results
Ability to predict the value of sales activities
Sales reps able to invest their time in an optimal way and avoid ineffective activities
Sales managers enabled to optimize their team size and predict how much revenue is generated by additional activities

Dig Deeper

Dig Deeper

Listen to Episode 30 of the Let’s Talk Data podcast and read the blog.

Scenario content provided by Syskoplan Reply

At a Glance

Sales Order
Completion

Accelerate the processing of complex sales orders

AT A GLANCE

A global chemicals company that creates 15 million sales order items yearly wanted to streamline order processing. High volumes of sales documents originated from multiple locations. Yet missing information and manual processing delayed order fulfillment, compromising customer satisfaction.

Challenges

Challenges

Incomplete sales orders, requiring significant manual effort to finalize (see Figure)
Incorrect orders, forcing workers to check historic order data from customers before processing
Delayed order fulfilment, increasing the need to address customer service expectations

Solution

Solution

Data attribute recommendation services, based on AI functionality (see Figure 56)
Data integration with SAP ERP and SAP HANA
Creation of machine learning models using Jupyter and Python

Benefits

Benefits

Decreased manual effort needed to complete sales orders
Faster, more accurate sales order processing
Increased sales revenue, thanks to higher process efficiency

Dig Deeper

Dig Deeper

Read this blog to learn how SAP Data Intelligence can act as a machine learning operations platform.

At a Glance

Serial Returns Detection

Reduce product returns and protect profit margins

AT A GLANCE

A large retailer found that more than half of certain products were returned after purchase, and that some customers were much more likely to return goods than others. To incentivize customers to keep the merchandise they order and reduce the negative financial effects of returned goods, the company wanted to better understand return patterns (see Figure).

Challenges

Challenges

Lack of insight into which products were likely to be returned
Inability to predict the predisposition of customers to return merchandise
Reduced cash, profits, and margin percentage caused by returns
Increased cost of restocking labor, markdowns, out-of-stocks, and other expenses

Solution

Solution

Powerful data platform that supports consumption of data from multiple sources (see Figure)
Flexibility that allows data scientists to use their preferred language and tools to develop schemas

Rapid proof-of-concept development
Kubernetes technology, including components for orchestration, scheduling, and monitoring

Benefits

Benefits

Reduced rate of returns by changing customer incentives in the shopping basket
Decreased financial loss to the company from returned goods
Improved management of stock and product reordering
Significant savings on shipping cost for retailers and manufacturers
Enhanced understanding of true sales revenue

Dig Deeper

Dig Deeper

Create your first machine learning scenario using the SAP Data Intelligence Cloud solution.

At a Glance

Service Ticket Intelligence

Automate service processes for improved customer support

AT A GLANCE

To maintain its reputation as a trusted brand, a manufacturer of home security products wanted to reduce the response time for customer service tickets. The company needed to streamline its support processes and deploy a more efficient way of categorizing incoming service requests (see Figure).

Challenges

Challenges

Incomplete or imprecise problem descriptions and ticket categorization
Complex service category hierarchy
Repetitive ticket categorization tasks with little added value

Solution

Solution

Embedded ticket intelligence capabilities used to automate classification tasks (see Figure)
Multilingual, natural-language understanding of service requests, which can be submitted using e-mails, tweets, phone calls, or text messages

Integration of ticketing, customer support, and back-end business systems

Benefits

Benefits

Accelerated process of addressing and resolving customer issues
Reduction in first-response times
Increased customer satisfaction
Empowerment of service representatives with recommendations based on historical data
Decreased time spent on manual tasks, such as sorting e-mails

Dig Deeper

Dig Deeper

Listen to Episode 37 of the Let’s Talk Data podcast: Cognitive Customer Service.

At a Glance

Time Sheet Analytics

Understand external workforce spend

AT A GLANCE

The IT department of a major corporation needed to increase its ability to understand and control spending on contingent workers. Although the company collected time-sheet data in one cloud-based system and business-partner data in an on-premise enterprise software solution, it struggled to create one unified view of the information (see Figure).

Challenges

Challenges

Huge volumes of diverse data on contingent workers, stored in different formats
Inability to integrate workforce and business-partner data
Different sources of data (cloud-based and on-premise solutions)
No view of trends across various global offices, impeding the ability to improve hiring and maximize use of contingent workers

Solution

Solution

Integration among various data sources using open application programming interfaces (see Figure)
Data masking and preprocessing functionality that aligns data, creating common data formats

Data warehouse that calculates aggregated business metrics
Advanced time-sheet analytics dashboard for intuitive user access to workforce insights

Benefits

Benefits

Greater insight into external workforce spend, enhancing decision-making
Improved data consolidation and management of information about contingent workforce spend
Reduced human error in spend data, improving data quality

Dig Deeper

Dig Deeper

Learn more about data integration.

At a Glance

Automated Supplier Invoice Management

Eliminate duplicate invoices to better control spend

AT A GLANCE

Duplicate invoices should be recognized as part of any audit analytics program to ascertain whether the company being invoiced is paying for services or items for which they have already paid. Duplicate invoices can result from manual or processing errors. Regardless, recognizing duplicate invoices is critical to eliminating spend leakage.

Challenges

Challenges

Manual invoice data input that can result in duplicate entries for the same purchase invoice or payment request
Application limitations that are intrinsically constrained in terms of case types, formats, and the use of special characters – all associated with manual, error-prone entry of invoice numbers
Formatting inconsistencies that make it difficult to identify duplicate invoices, resulting in possible revenue loss

Solution

Solution

Automated microservice for invoice validation that identifies duplicates, reducing manual labor and avoiding cash leakage
Accurate detection of duplicates between payables and supplier invoices
Streamlined and automated reporting across multiple applications

Stronger data security through the use of encryption keys
Infrastructure that is scalable, robust, and secure for future use cases and validations

Benefits

Benefits

Reduction in spend leakage, thanks to automatic duplicate invoice detection
Flexible microservice technology that is fast and simple to deploy
Streamlined and automated reporting
Stronger data security through the use of encryption keys
Infrastructure that is scalable, robust, and secure for future use cases and validations

Dig Deeper

Dig Deeper

Read the blog to learn more.

Scenario content provided by DXC Technology

At a Glance

Commodity Price Prediction

Stabilize spend in a fluctuating economy

AT A GLANCE

Organizations constantly strive to cut their bottom-line expenses, specifically their procurement costs related to raw materials that are the key ingredients to finished goods. Commodities like oil, copper, gold, grains, and so on have a major impact on the global economy. As such, there is a big demand for foresight into fluctuations in pricing. Companies owning key data and midlevel technology capabilities need a solution that enables mission-critical decision-making, while significantly reducing the cost and time for making decisions.

Challenges

Challenges

Variety of systems to be integrated
Difficulty finding relations between different data sources
Missing data, making it difficult to use predictions

Lack of modelling interface
Decision trail that is difficult to establish
Collaboration that is challenging

Solution

Solution

Integration with various sources, such as SAP software, databases, data warehouses, files, APIs, weather, transportation issues
Outlier analysis and correction
Modelling, pipeline building, and automation

Commodity Price prediction
Visualization, simulation, and collaborative decisioning

Benefits

Benefits

Automated work flow
Performance improvement with data volumes
Control of costs and data models
Visibility into future spend
Historical data comparisons
Higher margins and yield
Lower costs
Optimized workforce
Integrated IT architecture
Simple solution deployment and maintenance

Dig Deeper

Dig Deeper

Read the blog to learn more.

Scenario content provided by VASPP

At a Glance

Maintenance and Reliability Intelligence

Improve the efficiency of asset maintenance with integrated analysis of maintenance execution processes

AT A GLANCE

Process improvement is critical to improving asset maintenance. This solution provides a point of view for management to explore process efficiency and track common key performance indicators (KPIs) for asset maintenance. Organizations also get the system robustness and flexibility they are looking for. The result is improved maintenance processes and capacity usage as well as greater maintenance plan accuracy.

Challenges

Challenges

Generic asset maintenance KPIs
Inflexible asset tracking mechanisms
Lack of visibility into end-to-end maintenance execution processes
Inability to update plans and pivot in the instance of delays or issues
No impact simulation or assessment ability

Solution

Solution

Rich dashboard reporting to provide insights on various facets of important asset maintenance KPIs
Spatial analysis and map visualization
Historical snapshot to support trend analysis

Process cockpit that integrates SAP and third-party solution workflows to provide a 360-degree view of maintenance execution processes
Machine learning to predict subsequent process steps and simulate changes in assumptions

Benefits

Benefits

Optimized maintenance capacity usage and maintenance processes as well as greater plan accuracy
Improvement in asset usage
Ability to identify opportunities for improvement, including where robotic process automation can help solve inefficiencies
Ability to identify bottlenecks and improve process and capacity utilization for better visibility and control
Enhanced contingency planning through work order process simulation

Dig Deeper

Dig Deeper

Read the blog to learn more.

Scenario content provided by DXC Technology

At a Glance

Robotic Process Automation for Quote and Order Processing

Automate sales processes to scale efficiently

AT A GLANCE

A wholesale distributor in the oil and gas industry needed to address the challenges of processing large volumes of quotes and orders. Specific challenges included improving accuracy and reducing the time to create quotes and orders. Leveraging the power of machine learning, the company wanted to automate quote and order processing to compete for business with greater efficiency and speed. Robotic process automation (RPA) paired with proprietary search algorithms was used to identify material and customer data, from customer requests to automated sales document creation. The RPA feedback loop improved results and accuracy over time. The use of RPA within sales processes can automate time-consuming steps and improve sales process accuracy and win rates.

Challenges

Challenges

Incongruent data
Manual cross-referencing and data entry
Slow processing of sales quotes and orders
Dependence on human experience, memory, and performance of customer service representative (CSR)

Incongruent data between customer, vendor, and internal systems that creates data-matching challenges
Delayed sales team document turnaround that impacts the ability to win deals, leading to lost bids, lower sales, and lower service levels

Solution

Solution

RPA that makes data entry using incongruent data quick and accurate, allowing sales team to focus on selling to the customer
Machine learning cross-referencing that matches data with search-engine-type algorithms

Predictions that learn from training and feedback to continuously improve results
Scalable sales tools that make winning bids automatic, increasing sales potential
Faster, more reliable service that builds trust

Benefits

Benefits

Scale sales with automation
No SAP solution experience required
Improve customer buying experience
Allow customer service representatives to focus on customer relationship management and customer success
Reduce repetitive work and human error
Simplify quote and order creation
Reduce time to respond to customers and improve time to close

Dig Deeper

Dig Deeper

Read the detailed solution summary and watch the video.

Scenario content provided by DataXstream

At a Glance

Supply Chain Resilience Cockpit

Experience guided decision-making in SAP S/4HANA

AT A GLANCE

In today’s fast and volatile world, competitive supply chains are global and complex. Supply chain resilience is critical in a connected, cloud-based world. The Camelot Supply Chain Resilience Cockpit application uses the SAP Data Intelligence solution and the SAP Data Warehouse Cloud solution to capture and analyze data on external factors that are impacting supply chains. It then brings that external data together with internal data to help achieve greater transparency. Integrated with SAP S/4HANA, it enables guided decision-making based on historical and current data points. By using the power of data from multiple sources to identify risks in the supply chain, the solution makes it possible to react in real time to make informed decisions about rerouting or using alternative sourcing.

Challenges

Challenges

High volatility of supply chains due to difficulties with inbound and outbound deliveries
No reliable delivery forecasts possible
No transparency of risks due to external factors such as local events or weather

Solution

Solution

Integration with core ERP system landscape, such as SAP S/4HANA
Information about potential delays before order placement as well as the proposal of alternative options
Regional route and sentiment analysis that takes current, historical, and external data into account

Geovisualization of supply chains, including external data and alerts
Risk analysis and classification of suppliers when placing an order in the ERP system
Identification of trends and patterns based on suppliers, plants, and routes to calculate delay probabilities

Benefits

Benefits

Enhanced transparency into end-to-end supply chain processes through tight integration with ordering in SAP S/4HANA
Optimized prediction of major supply chain disruptions
Increased efficiency and smoother supply chain process execution
Consideration of relevant data for decision-making
Potential integration to further operational systems
Reduction in process disruptions and increase in on-time deliveries, thanks to alternative sourcing proposals

Dig Deeper

Dig Deeper

Listen to Episode 20 of the Let's Talk Data podcast: The Role of DataOps in Supply Chain.

Scenario content provided by Camelot ITLab GmbH

	Maximize the value of your data – by discovering hidden patterns and correlations for new insights and understanding
	Operationalize and scale machine learning – providing data that is fit for use and ensuring machine learning can be scaled across the enterprise
	Simplify and unify data – so the right information gets into the right hands, at the right time, and in the right context to make critical decisions

	Discover and connect to any data, anywhere, and anyway you need
	Integrate and orchestrate massive data volumes and streams at scale
	Streamline, operationalize, and govern machine learning driven innovation

	Discovery and cataloging of distributed metadata, enabling a searchable data fabric
	Profiling and enhancing your data in order to drive data value
	Continuous monitoring to ensure robust data quality

	Centrally manage and process data wherever it resides
	Automate and reuse on-premise and cloud processing engines
	Manage complex data types across distributed environments


	Transform valuable data from any source into critical insight to drive better business outcomes.

Please rotate your device