Navigating Enterprise AI: Swisscom’s Journey with Amazon Bedrock AgentCore

How Swisscom is Leading the Charge in Scalable, Sustainable AI Solutions

Navigating the AI Ecosystem: Swisscom’s Approach to Scalable AI Solutions

This post was written with Arun Sittampalam and Maxime Darcot from Swisscom.

As we navigate the constantly shifting AI ecosystem, enterprises face significant challenges in translating AI’s vast potential into scalable, production-ready solutions. Swisscom, Switzerland’s leading telecommunications provider, exemplifies how organizations can navigate this complexity successfully while maintaining their commitment to sustainability and excellence, boasting an estimated revenue of $19B by 2025 and a market capitalization exceeding $37B as of June 2025.

A Commitment to Sustainability and Innovation

Recognized as the Most Sustainable Company in the Telecom industry for three consecutive years by World Finance magazine, Swisscom has established itself as an innovation leader committed to achieving net-zero greenhouse gas emissions by 2035, aligning with the Paris Climate Agreement. This sustainability-first approach extends to their AI strategy, as they break through what they call the “automation ceiling” — the limitations of traditional automation approaches that fail to meet modern business demands.

In this post, we explore how Swisscom implemented Amazon Bedrock AgentCore to build and scale enterprise AI agents for customer support and sales operations. As an early adopter of Amazon Bedrock in the AWS Europe Region (Zurich), Swisscom leads in enterprise AI implementation, employing their Chatbot Builder system and various AI initiatives, including Conversational AI powered by Rasa and fine-tuned LLMs on Amazon SageMaker.

Swisscom’s Agentic AI Enabler Framework

The challenge of scaling AI agents enterprise-wide lies in managing siloed agentic solutions while facilitating cross-departmental coordination. Swisscom addresses this through Model Context Protocol (MCP) servers and the Agent2Agent protocol (A2A), ensuring seamless communication across domains. Operating under Switzerland’s strict data protection laws, they developed a framework balancing compliance requirements with efficient scaling capabilities, helping prevent redundancies while maintaining high-security standards.

Challenges in Multi-Agent Architecture

Swisscom’s vision for enterprise-level agentic AI confronts fundamental challenges that organizations face when scaling up AI solutions. Successful implementation requires more than just innovative technology; it demands a comprehensive approach to infrastructure and operations. Among these challenges is orchestrating AI agents across different departments while ensuring security and efficiency.

Consider a common customer service scenario where an agent is tasked with helping a customer restore Internet router connectivity. The issue could stem from a billing issue, a network outage, or a configuration mismatch, each residing in different operational domains. This exemplifies the need for seamless cross-departmental coordination.

Addressing the Challenges With Amazon Bedrock AgentCore

Amazon Bedrock AgentCore equips Swisscom with a comprehensive solution, addressing enterprise-scale agentic AI challenges through various key features:

1. AgentCore Runtime: By enabling Swisscom’s developers to build agents efficiently while ensuring secure and cost-effective hosting, the system handles scaling through Docker container deployments that preserve session-level isolation.

2. AgentCore Identity: This feature integrates smoothly with Swisscom’s identity provider, managing authentication and authorization without the complexities of custom token exchange servers.

3. AgentCore Memory: This robust solution manages both session-based and long-term memory storage, vital for understanding customer interactions. It ensures that user data remains separate, supporting heightened security and compliance efforts.

4. Strands Agents Framework: With its simplified agent construction and rapid development cycles, it has gained high adoption among Swisscom’s developers, facilitating seamless integration with Bedrock AgentCore services.

Implementing Real-World Solutions

Swisscom partnered with AWS to implement Amazon Bedrock AgentCore for two B2C cases: generating personalized sales pitches and providing automated customer support for technical issues. These agents are integrated into Swisscom’s existing customer-gen AI-powered chatbot system, SAM, which requires high-performance agent-to-agent communication protocols.

This implementation has proven beneficial:

• Development teams achieved stakeholder demos within 3-4 weeks.
• Migration to the Strands Agents framework reduced complexity and accelerated development cycles.
• The framework’s native OpenTelemetry integration helped maintain consistency with enterprise-wide monitoring standards.

Key Insights and Strategic Implications

Swisscom’s implementation of Amazon Bedrock AgentCore offers valuable insights for enterprises looking to navigate the complexities of productive, scalable AI:

1. Architectural Foundation Matters: Addressing secure authentication and agent orchestration challenges has established a scalable foundation that accelerates deployment.

2. Framework Selection Drives Velocity: The right development tools can dramatically reduce time-to-value; agent teams achieving significant milestones in weeks validate this principle.

3. Compliance as an Enabler: Swisscom’s success illustrates that regulatory compliance can coexist with innovation, proving that scaling while maintaining data sovereignty is possible.

Looking Ahead: Continuous Improvement and Governance

The future roadmap focuses on agent sharing, cross-domain integration, and governance. A centralized agent registry will facilitate resource discovery and reuse while enabling seamless collaboration between different business units. Robust governance mechanisms, including version control and regular audits, will support sustainable growth while maintaining compliance with enterprise standards.

Conclusion: A Blueprint for Success

Swisscom’s journey through the complexities of production-ready Agentic AI serves as a blueprint for organizations aiming to deploy scalable AI solutions effectively. Their success with high-volume B2C applications illustrates that agentic AI not only delivers measurable business outcomes but can also thrive on a solid architectural foundation. By leveraging thoughtful planning and innovative technology choices, enterprises can tap into the transformative potential of AI in customer service and sales operations.

By staying ahead of industry trends and prioritizing compliance, Swisscom is not only enhancing its operational capabilities but also setting a precedent for future enterprise AI implementations. As organizations strive to leverage AI for competitive advantage, Swisscom’s experience serves as a beacon of innovative practice and strategic foresight.