Unlocking the power of generative AI with Instana and AWS

As the adoption of generative AI (GenAI) continues to surge, enterprises will need observability solutions that not only keep up with the demands of these complex applications but also empower them to make informed, data-driven decisions.  At AWS re:Invent this year, we are thrilled to showcase the evolution of our partnership with AWS and to announce a new product feature: Instana GenAI Observability for AWS Bedrock, available for public preview starting 2 December 2024.

Our collaboration with AWS has always been rooted in the shared goal of delivering high-performance, resilient, and cost-effective solutions. With over 20 native integrations for AWS services and plans to expand to more than 90, Instana helps provide deep visibility into AWS applications, helping organizations achieve end-to-end observability across their cloud ecosystems.

Earlier this year, we introduced GenAI Observability in Instana, designed for enterprises to monitor the performance of large language models (LLMs) and help support their value contribution to business objectives. Today, we are expanding this capability specifically to AWS Bedrock, allowing our AWS customers to monitor generative AI applications and their underlying LLMs in near real time.

The rise of GenAI, with its transformative impact across industries, has revealed new challenges in monitoring model performance and understanding its business impact. With this new feature, enterprises using AWS Bedrock can monitor their LLMs and associated applications, to help support an alignmentwith business goals. Our AI-powered insights offer observability, providing visibility into LLM performance, impact, and opportunities for optimization.

• Performance assurance: Instana can monitor model response times, error rates, and reliability with metrics such as model latency and API request counts to offer near real-time insights to deliver consistent performance.
• Resource optimization: By trying to track token usage, cost, and resource utilization, businesses can streamline operations, balance workloads, and control costs effectively.
• Anomaly detection: Instana’s intelligent alerts, supported by distributed tracing of AWS Bedrock API calls, can flag irregularities in model behavior and pinpoint performance bottlenecks to address challenges.
• Scalability and load balancing: Through visibility into API requests and system demand, Instana can help support dynamic GenAI workloads scale without compromising performance.

Instana’s GenAI Observability for Bedrock brings us closer to our vision of enabling cloud-native enterprises to maintain high-performance AI applications without sacrificing control over operational metrics.

Looking ahead to 2025, we plan to deepen our observability capabilities for AWS with plans to support AWS SageMaker. This will allow customers to leverage Instana's insights to monitor the end-to-end lifecycle of machine learning models, from training to deployment. With Instana’s ongoing contributions to OpenTelemetry and industry collaboration, we’re also exploring ways to enhance data translation for AI observability and provide seamless integration across observability tools.

Statements regarding IBM's future direction and intent are subject to change or withdrawal without notice and represent goals and objectives only.

Sign up for the public preview.

Learn more about IBM Instana LLM Observability.