Driving Supply Chain Operations at Speed with IBM Consulting Supply Chain Ensemble on AWS

By Anu Singh, Global Supply Chain Transformation Leader – IBM Consulting
By Jay Pandya, Partner Solutions Architect – AWS

IBM

In today’s dynamic business environment, organizations face numerous challenges in managing their high-volume and high-speed supply chains efficiently. However, conventional supply chain management applications often fall short in addressing the complexities of achieving end-to-end visibility.

Meanwhile, the recent disruptions from prolonged lockdowns and demand fluctuations during the pandemic further exposed the lack of enhanced interconnectivity and real-time data accessibility among supply chain partners.

While the concept of “control towers” has been explored as a solution, their capabilities may not suffice for the process and data orchestration demands of today’s scale. This calls for a custom-built platform tailored for specific needs of organizations, steering away from the traditional notions of control towers.

Recognizing this need, IBM Consulting and Amazon Web Services (AWS) have collaborated to co-create a supply chain framework after analyzing and refining 100+ use cases, focusing on risk identification and management of exceptions.

In this post, we will explore how IBM Consulting Supply Chain Ensemble on AWS helps customers overcome common challenges and deliver efficiency as well as optimization within their supply chains.

IBM Consulting is an AWS Premier Tier Services Partner that helps customers who use AWS to harness the power of innovation and drive their business transformation.

Ensemble: High Speed, Agile, Flexible Supply Chain Platform

IBM has collaborated with AWS to create a no-license, hyperscaler-based supply chain framework. IBM Consulting Supply Chain Ensemble on AWS is a comprehensive and innovative framework designed to enable custom-built solutions that go beyond the limitations of traditional methods, utilizing ready-to-use components.

Ensemble is a powerful combination of IBM’s data platform experience, artificial intelligence (AI) capabilities, and supply chain expertise, integrated with AWS services and best practices.

The framework offers modularity, enabling organizations to expedite implementation cycles and make swift progress. It leverages the concept of Digital Twin+ to create near real-time representations of key business processes, enabling end-to-end visibility and faster planning and response.

By infusing ecosystem-centricity and advanced analytics into the solution, Ensemble is built on three key features that enhance its capabilities:

• Next best actions to manage exceptions: Ensemble recommends optimal actions using pre-configured business rules and collaborative workflows to mitigate risks, resolve discrepancies, manage incidents, and detect demand variations. Its real-time decisioning empowers organizations to make informed choices supported by intelligent analytics which enable root cause analysis.
• Collaborative actions for improved risk mitigation: It’s a comprehensive solution that combines planning, predicting, and mitigating supply chain risks to enhance resiliency and visibility. It fosters collaborative decision-making among supply chain partners, promoting customer-centricity and overall resiliency.
• End-to-end visibility for supply chain resilience: Ensemble guarantees comprehensive visibility throughout supply chain networks and identifying impacts across all functions. With horizontal and vertical traceability, it enables multi-tier visibility of inventory, purchase orders, suppliers, and production capacity.

Figure 1 – Ensemble has modular capabilities across supply chain domains.

Key Features

Businesses can leverage this framework to build their own use cases supported by the following features:

• Supports a common network data model with over 1,000+ data entities and attributes, designed to support 100+ supply chain use cases.
• Creates a high-speed platform through data ingestion with Kafka analytics and through AI capabilities built under Cognitive Classifier, which enables a faster source to destination data mapping.
• Real-time data integration through APIs, connecting enterprise and partner systems while incorporating traditional and telemetric data.
• Facilitates process orchestration for seamless ecosystem collaboration and transformative workflow enhancements.
• Empowers intelligent decision making through generative AI and highly accurate prescriptive and predictive machine learning (ML) models.
• Embraces an agile and flexible architecture with configurable components, offering a plug-and-play approach that allows organizations to leverage their existing investments seamlessly.

One Network Data Model

Ensemble offers a ready-to-use network data model with 1,000+ entities and attributes, supporting popular enterprise applications like SAP, EDI, MFST CDM, and more. It ensures compliance through GS1 and OAGi standards and accommodates various data sources.

Powered by Agent-Skill architecture, it provides out-of-the-box connectors for standard enterprise resource planning (ERP) and external datasets. Ensemble enables seamless integration of Internet of Things (IoT) data, enhancing inventory visibility, rotation compliance, and stock recovery. Ensemble’s data platform architecture is scalable to support a billion+ transactions.

High-Speed Network Platform

Ensemble enables creation of high-speed data ingestion through Kafka analytics followed by intelligent data classification, mapping, and anomaly detection. The framework utilizes AI to expedite mapping of source and destination data at scale through Cognitive Classifier. An intelligent feedback system allows users to review and improve data records. Integrated with Spark, it supports model training and prediction, leveraging user feedback to enhance accuracy.

Generative AI

Training with new sets of data over and above the traditional training methods to generate supply chain specific solutions. The Ensemble platform uses generative AI techniques based on large language models (LLMs) to answer questions in our data, generate communication templates between key stakeholders and identify discrepancies and recommend actions to resolve them.

AI Capabilities

Ensemble offers a collaborative AI framework that enables multi-agent collaboration, customization, forecasting capabilities, hybrid cloud support, and 24/7 availability. Demand Sensing Advisor utilizes collaborative agents to assess and select the best demand forecasting model. Continuous intelligent planning incorporates industry-trained AI algorithms to improve forecast accuracy and optimize supply chain operations.

The Sales & Operations Planning feature integrates the annual operating plan with demand and supply planning, while the What-if Analysis helps achieve the right balance between inventory investment and service levels. Advanced Forecasting considers over 70 external influencing variables using Multiple Linear Regression Model (MLR) and ARIMAX – Auto-Regressive Integrated Moving Average.

Lastly, AI at Scale is a service that facilitates the integration, scaling, and management of ML models into production environments.

End-to-End Traceability

Blockchain technology enables real-time visibility and tracking of goods across the supply chain, enhancing traceability and combating issues like counterfeiting, compliance violations, and delays. Transactions are automatically updated, providing accurate and time-stamped information. Ensemble combines blockchain with AI and IoT data to drive next best actions.

Figure 2 – Consolidated map view showing discrepancies/delays.

Flexibility and Reusability

Ensemble’s adaptable architecture and plug-and-play approach seamlessly integrates with existing systems and processes, addressing the distinct needs of each organization. This flexibility allows organizations to capitalize on their current investments and infrastructure, and expedites the implementation.

By harnessing reusable components and a diverse range of ready-to-implement use cases, you can gain a substantial advantage in project initiation, avoiding the complexities of building from scratch.

Figure 3 – Development benefits.

Hyper-Personalization

Ensemble is at the forefront of the changing landscape in how we interact. Embracing generative AI, IBM’s system facilitates seamless and personalized communication between users and technology. This transformation is revolutionizing the way we engage with information and make critical decisions in the supply chain, optimizing operations for a more customized and efficient future.

Figure 4 – Personalized Conversational Interface.

Multi-Tier Supplier View

Ensemble offers a unique and powerful feature—a multi-tier view of suppliers which incorporates network data to reveal the complex web of relationships between entities. This feature provides a bird’s-eye perspective, enabling organizations to visualize supplier networks in a clear and intuitive manner.

By mapping out these connections, businesses can improve their plans, enhance collaboration, and strengthen their supply chain resilience.

Figure 5 – Multi-tier supplier view.

Reference Architecture on AWS

IBM collaborated with AWS and its 200+ service offerings to combine technology services and scalability to develop a supply chain framework which addresses, resolves, and enhances challenges across multiple industries.

Figure 6 – Supply chain data platform architecture on AWS.

Modular, Scalable, On-Demand and Cost Efficient Portfolio

To support a high-speed, high-volume application, Ensemble leverages several key AWS services

• Real-time data pipelines and streaming.
• AWS serverless computing and container-based microservices for distributed and scalable processing.
• Data lake on AWS for centralized storage and processing of structured and unstructured data.
• Amazon QuickSight for interactive dashboards and visualizations.
• Amazon DynamoDB for low-latency and high-throughput NoSQL database operations.
• Amazon API Gateway for managing APIs at scale.

These services and others provide the foundation for a high-performance application with capabilities for data storage, processing, analytics, visualization, and API integration.

Figure 7 – AWS components architecture for Ensemble.

Ensemble leverages the following AWS services in building the platform:

Data Storage and Management

• Integration partners generate source data in three formats: batch files, streaming IoT/RFID data, and published APIs for consumption by Ensemble. Input files are stored in an Amazon Simple Storage Service (Amazon S3) bucket.
• Amazon Managed Streaming for Apache Kafka (Amazon MSK) connectors registered with the bucket monitor incoming files and trigger processing when new files arrive
• Amazon S3 acts as a data lake repository to store both raw and curated data that flows from the source input through Amazon MSK.
• Amazon DynamoDB serves as the database for the AI sandbox. Another instance is set up for blockchain-related tasks.
• Amazon Aurora PostgreSQL is a relational database management system (RDBMS) that hosts the relational data needed for consumption by services or applications.
• AWS Glue Crawler and AWS Glue Data Catalog are utilized for scanning databases and collecting metadata to build a data catalog as part of the data governance framework.

Data Processing and Analytics

• Amazon MSK serves as the central ingestion point for the supply chain data platform. It receives data from the source input through connectors hosted on Amazon Elastic Kubernetes Service (Amazon EKS). MSK acts as a message broker, forwarding messages to the next step for additional processing tasks such as filtering and mapping. Confluent Kafka may be utilized if schema evolution/registry is required.
• Additional data processing, orchestration tasks, and business/application logic are implemented using AWS Lambda functions or microservices.
• AI framework provides capabilities for developing, training, managing, and deploying machine learning models. It integrates with various AWS AI tools, including Amazon SageMaker (analytics), Amazon Forecast (time-series), and Amazon Neptune (GraphDB), enabling a wide range of use cases.

Visualization and Reporting

• Amazon QuickSight serves as a dashboard for monitoring the status of services and operations.

API and Data Integration

• AWS Fargate hosts microservices that serve data to the APIs.
• Amazon API Gateway routes API requests to internal endpoints representing the microservices hosted on EKS.

Conclusion

Organizations can leverage IBM Consulting Supply Chain Ensemble on AWS to build a modular and scalable platform centered around their supply chain ambitions.

Ensemble can be leveraged by organizations to build a custom platform that will enable you to swiftly respond to and manage any deviations or disruptions in the supply chain. It also enables you to collaborate effectively with business partners and harness intelligence to inform strategic decisions made by supply chain users across the network.

Read more about IBM Consulting capabilities on AWS.

Special thanks to our contributors and collaborators: Ryan Keough, AWS; Nisha Dekhtawala, AWS; Rajkiran Veluri, IBM Consulting; Melvin Tomy, IBM Consulting; Amit Jindal, IBM Consulting; and Akshay Bhambure, IBM Consulting.

.

.

IBM – AWS Partner Spotlight

IBM Consulting is an AWS Premier Tier Services Partner and MSP that offers comprehensive service capabilities addressing both business and technology challenges that clients face today.

Contact IBM | Partner Overview | Case Studies