Cisco uses Amazon Bedrock as the foundation for delivering generative AI features across its portfolio. Bedrock gives Cisco easy access to a range of cutting-edge models from multiple providers through a single managed service. Because Bedrock is built on AWS’s global cloud infrastructure, Cisco can: - Deploy AI features in many regions to be closer to customers - Keep latency low for end users - Avoid large upfront infrastructure investments This setup helps Cisco scale AI features as demand grows, while keeping operations cost-effective and manageable. By building on Bedrock, Cisco can focus more on designing useful AI experiences and less on running and maintaining the underlying AI infrastructure.

For Cisco and AWS, responsible AI is about designing and operating generative AI systems in a way that builds and maintains customer trust. In this context, responsible AI focuses on three main principles: 1. **Fairness** – Working to reduce bias in AI outputs so that results are as fair and consistent as possible across different users and use cases. 2. **Transparency** – Being clear about how AI features are built and used, and helping customers understand what AI is doing behind the scenes. 3. **Security** – Protecting data, models, and AI-powered applications using AWS’s security capabilities and Cisco’s own security practices. By combining these principles with Amazon Bedrock’s managed environment, Cisco aims to reimagine how AI is delivered—making it useful and scalable while still aligning with governance, compliance, and trust requirements.

Cisco chose Amazon Bedrock because it provides a practical balance of flexibility, scale, and cost control for generative AI: - **Access to multiple leading models**: Bedrock offers a selection of advanced models from different providers through one service, so Cisco can pick the right model for each use case without integrating each provider separately. - **Global scalability and low latency**: Built on AWS’s global cloud, Bedrock lets Cisco deploy AI features closer to users in many regions, which helps reduce latency and improve user experience. - **Cost-effective scaling**: Instead of making large upfront investments in specialized AI infrastructure, Cisco can scale usage up or down with demand, paying for what it needs as it grows. - **Security and managed operations**: Bedrock is a fully managed service, which reduces the operational burden of running and securing AI infrastructure. This allows Cisco’s teams to focus on innovation and product features rather than infrastructure management. Overall, Bedrock helps Cisco rethink how it brings AI to market—speeding up delivery while staying aligned with its responsible AI and security commitments.