By Becky Carroll

AI has the potential to unlock trillions in value over the next decade with generative AI at the forefront, reinventing experiences and creating never-seen-before applications. Issues around implementation, cost and trust can often be roadblocks to value, but those challenges can be addressed by approaching AI holistically with a clear strategy—from the data that trains it to the infrastructure it runs on to the problems it’s designed to solve.

For many companies, one of the challenges in realizing value from GenAI is an inability to quickly innovate where it matters and deploy solutions infused with GenAI at speed and scale. Here we share two case studies of enterprises that have implemented enterprise platforms to improve efficiencies and adopt GenAI at scale.

Case Study 1: Leading Media and Entertainment Distributor Greatly Improves IT Service Delivery

Background: For nearly three decades, this large American media and entertainment distributor has been at the forefront of delivering innovative entertainment experiences to consumers. As a leader in the industry, the company has consistently adapted to changing consumer preferences, providing content in the ways that their audiences prefer.

Business and Technology Challenges: The client faced significant challenges in transforming user journeys across various business domains, including sales, ordering, provisioning, and billing. Key issues included:

1. Lack of business value generation and realization
2. Need for improved platform stability, security, resiliency, and maintainability
3. Suboptimal developer productivity due to inadequate platform support
4. Inefficient knowledge management framework
5. Time losses during critical phases of application development and deployment
6. Absence of platform capabilities to enable enterprise-wide Generative AI use cases
7. Inability to innovate quickly and cost-effectively in high-impact areas

The client sought a framework to rapidly develop Generative AI proofs of concept (POCs) and infuse GenAI into each phase of IT service delivery, with the ability to deploy solutions at speed and scale.

Solution: To address these challenges, the client partnered with IBM Consulting and AWS to build a comprehensive Generative AI platform leveraging Amazon Bedrock. The solution comprised the following key components:

1. GenAI Platform: A badge-less scrum team consisting of IBM Consulting, AWS, and client personnel collaborated to design and implement a robust GenAI platform architecture. This architecture leveraged Amazon Bedrock's foundation models (FMs) from leading AI companies, providing a flexible and scalable base for various GenAI applications.
2. LLM Governance Framework: The team established a Large Language Model (LLM) governance framework to ensure responsible and ethical use of GenAI technologies across the organization.
3. Foundational GenAI Capabilities: The team implemented over 30 GenAI foundational capabilities in production.

4. AWS Cloud Platform Enhancements: The existing AWS cloud platform was expanded and optimized to support the new GenAI capabilities. Key improvements included:

a. Automation of upgrades and patches

b. Optimization of EKS clusters and volumes

c. Enhanced CloudWatch monitoring and alerting

d. Implementation of advanced security measures

Outcomes and Benefits: The implementation of this comprehensive GenAI solution resulted in significant improvements across various aspects of the client's IT service delivery:

1. Cost Savings:

• $3M+ in yearly savings on application development expenses
• Projected annual savings on cloud services costs based on implemented improvements

2. Operational Efficiency:

• 35% reduction in incident ticket creation
• 25% improvement in developer productivity

3. Security Enhancements:

• Remediation of thousands of security vulnerabilities within several months

4. Innovation Acceleration:

• Rapid development and deployment of GenAI-powered solutions across the enterprise

7. Platform Stability and Resilience:

• Increased platform stability, security, and maintainability
• Enhanced developer support and productivity through optimized AWS cloud infrastructure

This case study demonstrates the transformative power of GenAI in revolutionizing IT service delivery. By leveraging these advanced GenAI technologies, the client has not only achieved significant cost savings and operational improvements but has also positioned itself at the forefront of innovation in the industry.

Case study 2: Global CPG Giant Streamlines Generative AI Adoption

Background: In the fiercely competitive consumer packaged goods (CPG) industry, innovation and efficiency are paramount. Our client, a leading global CPG company with a presence in over 180 countries, recognized the transformative potential of generative AI to revolutionize their operations and maintain their market leadership.

Business and Technology Challenges: The client faced a perfect storm of challenges as they sought to harness the power of generative AI:

1. Cost Management: Multiple overlapping use cases resulted in skyrocketing AI-related expenses.
2. Change Resistance: Employees struggled with adopting new technologies, hindering the introduction of flexible offerings.
3. Time-to-Market Pressure: The need to accelerate innovation while addressing change management created a challenging balancing act.
4. Proliferation of GenAI POCs: Multiple internal teams were independently creating generative AI proof-of-concept (POC) use cases, leading to a sprawling and unmanageable landscape of initiatives.
5. Resource Waste: Lack of integration between teams led to duplicate efforts and repetitive work.

Solution: To address these multifaceted challenges, the client partnered with IBM Consulting to develop a centralized generative AI platform leveraging Amazon Bedrock. This platform was designed for rapid onboarding and scaling of GenAI use cases in production. Key components of the solution include the following:

1. Amazon Bedrock Integration: The platform utilizes multiple models, using Amazon Bedrock to provide a centralized subscription management system for various LLM models and reduce the need for individual setups.
2. Reusable Components: The platform encapsulates reusable components of GenAI applications, offering them as API-based services. This includes:
• Data ingestion pipelines
• Vector database management
• Advanced search capabilities
• Retrieval Augmented Generation (RAG) frameworks
• Governance tools
• Integration with the Enterprise Data Platform
3. Amazon EKS Deployment: The solution leverages Amazon Elastic Kubernetes Service (EKS) for deploying applications with disaster recovery support, ensuring high availability and scalability.
4. Infrastructure as Code: The entire platform is fully automated using Terraform scripts, enabling consistent and repeatable deployments across the organization.
5. AI Governance Model: A centralized AI governance model was implemented to ensure responsible and ethical use of AI across all use cases.

Outcomes and Benefits:

The implementation of this centralized GenAI platform delivered significant benefits across the organization, including the following:

1. Streamlined AI Adoption: The company successfully controlled the proliferation of GenAI stacks, reducing the number of disparate AI initiatives.
2. Accelerated Time-to-Market: The platform reduced the average time to deploy new GenAI use cases and increased the ability to introduce flexible offerings to its customers.
3. Improved Governance: The centralized AI guardrails and governance model ensured compliance with corporate and regulatory requirements across GenAI initiatives.
4. Resource Efficiency: By promoting a "Do Not Repeat Yourself" (DRY) culture and leveraging reusable components, the company reduced duplicate development efforts.
5. Data Utilization: Integration with the Enterprise Data Platform led to an increase in the utilization of valuable corporate data assets in AI applications.
6. AI FinOps Integration: The platform implemented AIFinOps practices to optimize resource utilization and control expenses.

This case study demonstrates how the strategic implementation of a well-designed centralized GenAI platform can transform a large enterprise's approach to generative AI. By centralizing resources, streamlining processes, and fostering a culture of innovation, the company has positioned itself for long-term success in the rapidly evolving landscape of AI-driven innovation.

Discover the ways IBM Consulting and AWS can unleash the transformative value of generative AI in your business with greater speed, scale, and trust.

Read more: https://www.ibm.com/consulting/aws