Swisscom Accelerates Sales and Support: How Amazon Bedrock AI Agents Transformed Telecom

Swisscom, Switzerland's leading telecom provider, faced the challenge of scaling AI solutions amid strict data privacy requirements and the need for rapid adaptation. By implementing AI agents powered by Amazon Bedrock, they not only reduced development time to 3-4 weeks but also ensured scalability to handle thousands of requests per month in sales and customer support departments.

If your company is losing time on routine sales or support tasks and is looking for a scalable solution, message our manager — he will run a free analysis of your business and niche and show exactly how to get a real business result from an AI agent in your case, not a nice-looking picture. Message the manager

Challenges of a Telecom Giant: "Automation Ceiling" and Strict Regulations

Swisscom, a company with a turnover of approximately $19 billion, had long been involved in automation but reached an "automation ceiling" where traditional approaches no longer met modern business demands. The main problems were:

• Scaling. The complexity of coordinating AI agents across different departments while maintaining security.
• Security and Authentication. The need for secure, transitive access with least privilege, considering overlapping permissions (customer, agent, department).
• Integration. How to centralize and ensure compatibility with existing AI solutions and infrastructure.
• Analytics. Effective collection and use of customer data from multiple interactions.

For example, in the event of a customer's router failure, the problem could be billing, network, or configuration, each belonging to different departments. This required seamless cross-departmental coordination.

The Solution: AI Agents Powered by Amazon Bedrock AgentCore

Swisscom chose Amazon Bedrock AgentCore as the foundation for its AI agent architecture. This solution provided a comprehensive set of tools:

• AgentCore Runtime. Allows developers to focus on building agents, while the system handles secure hosting, automatic scaling, and session isolation.
• AgentCore Identity. Integrates with Swisscom's existing identity provider, managing inbound and outbound authentication. This simplifies secure interaction between agents, tools, and data sources.
• AgentCore Memory. Robust storage for both session-based and long-term memory, which is critical for B2C operations where understanding customer context across interactions is crucial.
• Strands Agents Framework. Simplifies agent creation, accelerates development cycles, and has built-in capabilities for tracing, evaluation, and logging.

Results and Benefits: From Concept to Real Product in Weeks

Swisscom implemented AI agents for two B2C scenarios: generating personalized sales pitches and automated customer support for technical issues. Both agents were integrated into Swisscom's existing SAM chatbot, requiring high performance and low latency.

Key results:

• Development Speed. Development teams, with no prior experience with Strands Agents, achieved their first business demos within 3-4 weeks.
• Reduced Complexity. One team migrated from LangGraph to Strands Agents, noting reduced complexity and accelerated development cycles.
• Scaling. AgentCore Runtime efficiently handles thousands of requests per month for each agent, maintaining low latency and optimizing costs.
• Regulatory Compliance. The system supports data sovereignty and user privacy, which is particularly valuable in Switzerland with its strict regulatory requirements.

Through this approach, Swisscom significantly accelerated AI adoption, improved sales and support efficiency, and ensured compliance with high security and privacy standards.

Source: aws.amazon.com

Want to learn how similar solutions can be implemented in your company? Message the manager. We will analyze your business for free and show the potential of AI agents.